Uploaded by midithru

Fundamentals of Electrical Engineering I 9648

advertisement
Fundamentals of Electrical
Engineering I
© Don Johnson
This work is licensed under a Creative Commons-ShareAlike 4.0 International License
Original source: The Orange Grove
http://florida.theorangegrove.org/og/items/c7d013ec-2b4d-6971-65f9-b1a6af2d9092/
1/
Contents
Chapter 1 Introduction ..................................................................................................1
1.1 Themes ................................................................................................................................1
1.2 Signals Represent Information ........................................................................................2
1.2.1 Analog Signals..........................................................................................................3
1.2.2 Digital Signals...........................................................................................................4
1.3 Structure of Communication Systems .............................................................................5
1.4 The Fundamental Signal ...................................................................................................7
1.4.1 The Sinusoid ............................................................................................................7
Exercise 1.4.1 ............................................................................................................8
Exercise 1.4.2 ............................................................................................................8
1.4.2 Communicating Information with Signals ............................................................8
1.5 Introduction Problems.......................................................................................................9
1.6 Solutions to Exercises in Chapter 1 ................................................................................10
Chapter 2 Signals and Systems ...................................................................................11
2.1 Complex Numbers ...........................................................................................................11
2.1.1 Definitions ..............................................................................................................11
Exercise 2.1.1 ..........................................................................................................13
Exercise 2.1.2 ..........................................................................................................13
2.1.2 Euler's Formula ......................................................................................................13
2.1.3 Calculating with Complex Numbers....................................................................14
Exercise 2.1.3
.....................................................................................................15
Example 2.1 ............................................................................................................16
2.2 Elemental Signals..............................................................................................................16
2.2.1 Sinusoids ................................................................................................................16
2.2.2 Complex Exponentials ..........................................................................................17
2.2.3 Real Exponentials ..................................................................................................18
2.2.4 Unit Step .................................................................................................................18
2.2.5 Pulse........................................................................................................................20
2.2.6 Square Wave ..........................................................................................................20
2.2.7 Signal Decomposition ...........................................................................................21
Example 2.2 ............................................................................................................21
Exercise 2.3.1 ..........................................................................................................21
2.3 Discrete-Time Signals ......................................................................................................21
2.3.1 Real-and Complex-valued Signals .......................................................................22
2.3.2 Complex Exponentials ..........................................................................................22
2.3.3 Sinusoids ................................................................................................................23
2.3.4 Unit Sample............................................................................................................23
2.3.5 Symbolic-valued Signals .......................................................................................24
2.3.6 Introduction to Systems .......................................................................................24
2.3.6.1 Cascade Interconnection...........................................................................25
2.3.7 Parallel Interconnection .......................................................................................25
2.3.8 Feedback Interconnection....................................................................................26
2.4 Simple Systems ................................................................................................................27
2.4.1 Sources ...................................................................................................................27
2.4.2 Amplifiers ...............................................................................................................27
2.4.3 Delay .......................................................................................................................28
2.4.4 Time Reversal.........................................................................................................28
Exercise 2.6.1 ..........................................................................................................28
2.4.5 Derivative Systems and Integrators ...................................................................29
2.4.6 Linear Systems.......................................................................................................29
2.4.7 Time-Invariant Systems ........................................................................................30
2.5 Signals and Systems Problems .......................................................................................32
Problem 2.1: Complex Number Arithmetic ................................................................32
Problem 2.2: Discovering Roots ...................................................................................32
Problem 2.3: Cool Exponentials ...................................................................................32
Problem 2.4: Complex-valued Signals .........................................................................33
Problem 2.5: ...................................................................................................................34
Problem 2.6: ...................................................................................................................35
Problem 2.7: Linear, Time-Invariant Systems .............................................................36
Problem 2.8: Linear Systems ........................................................................................37
Problem 2.9: Communication Channel .......................................................................37
Problem 2.10: Analog Computers ................................................................................38
2.6 Solutions to Exercises in Chapter 2 ................................................................................38
Chapter 3 Analog Signal Processing ...........................................................................40
3.1 Voltage, Current, and Generic Circuit Elements ...........................................................40
Exercise 3.1.1 ..................................................................................................................41
3.2 Ideal Circuit Elements .....................................................................................................41
3.2.1 Resistor ...................................................................................................................42
3.2.2 Capacitor ................................................................................................................42
3.2.3 Inductor ..................................................................................................................43
3.2.4 Sources ...................................................................................................................44
3.3 Ideal and Real-World Circuit Elements ..........................................................................44
3.4 Electric Circuits and Interconnection Laws....................................................................45
3.4.1 Kirchhof's Current Law .........................................................................................46
Exercise 3.4.1 ..........................................................................................................47
3.4.2 Kirchhof's Voltage Law (KVL) ................................................................................47
Exercise 3.4.2 ..........................................................................................................48
3.5 Power Dissipation in Resistor Circuits ...........................................................................48
Exercise 3.5.1 ..................................................................................................................50
Exercise 3.5.2 ..................................................................................................................50
3.6 Series and Parallel Circuits ..............................................................................................50
Exercise 3.6.1 ..................................................................................................................52
Exercise 3.6.2 ..................................................................................................................54
Example 3.1 ....................................................................................................................55
Exercise 3.6.3 ..................................................................................................................56
3.7 Equivalent Circuits: Resistors and Sources....................................................................57
Exercise 3.7.1 ..................................................................................................................59
Example 3.2 ....................................................................................................................59
Exercise 3.7.2 ..................................................................................................................61
3.8 Circuits with Capacitors and Inductors ..........................................................................62
3.9 The Impedance Concept..................................................................................................63
3.10 Time and Frequency Domains ......................................................................................65
Example 3.3 ....................................................................................................................67
Exercise 3.10.1 ................................................................................................................68
3.11 Power in the Frequency Domain ..................................................................................68
Exercise 3.11.1 ................................................................................................................69
Exercise 3.11.2 ................................................................................................................70
3.12 Equivalent Circuits: Impedances and Sources ............................................................70
Example 3.4 ....................................................................................................................72
3.13 Transfer Functions .........................................................................................................73
Exercise 3.13.1 ................................................................................................................76
3.14 Designing Transfer Functions .......................................................................................76
Example 3.5 ....................................................................................................................77
3.15 Formal Circuit Methods: Node Method .......................................................................79
Example 3.6: Node Method Example ..........................................................................82
Exercise 3.15.1 ................................................................................................................83
Exercise 3.15.2 ................................................................................................................84
3.16 Power Conservation in Circuits.....................................................................................84
3.17 Electronics ......................................................................................................................86
3.18 Dependent Sources........................................................................................................86
3.19 Operational Amplifiers...................................................................................................89
3.20 Inverting Amplifier..........................................................................................................90
3.21 Active Filters ....................................................................................................................91
Example 3.7 ....................................................................................................................92
Exercise 3.19.1 ................................................................................................................93
3.22 Intuitive Way of Solving Op-Amp Circuits ....................................................................93
Example 3.8 ....................................................................................................................95
3.23 The Diode ........................................................................................................................96
3.24 Analog Signal Processing Problems .............................................................................98
Problem 3.1: Simple Circuit Analysis ...........................................................................98
Problem 3.2: Solving Simple Circuits ...........................................................................99
Problem 3.3: Equivalent Resistance ............................................................................99
Problem 3.4: Superposition Principle ....................................................................... 100
Problem 3.5: Current and Voltage Divider ............................................................... 101
Problem 3.6: Thevenin and Mayer-Norton Equivalents ......................................... 101
Problem 3.7: Detective Work ..................................................................................... 102
Problem 3.8: Bridge Circuits ...................................................................................... 102
Problem 3.9: Cartesian to Polar Conversion ........................................................... 103
Problem 3.10: The Complex Plane ............................................................................ 103
Problem 3.11: Cool Curves ........................................................................................ 103
Problem 3.12: Trigonometric Identities and Complex Exponentials .................... 104
Figure 3.61Problem 3.13: Transfer Functions ......................................................... 104
Problem 3.14: Using Impedances ............................................................................. 105
Problem 3.15: Measurement Chaos ......................................................................... 105
Problem 3.16: Transfer Functions ............................................................................ 106
Problem 3.17: A Simple Circuit .................................................................................. 106
Problem 3.18: Circuit Design ..................................................................................... 107
Problem 3.19: Equivalent Circuits and Power ......................................................... 107
Problem 3.20: Power Transmission .......................................................................... 108
Problem 3.21: Optimal Power Transmission ........................................................... 109
Problem 3.22: Big is Beautiful ................................................................................... 110
Problem 3.23: Sharing a Channel ............................................................................. 110
Problem 3.24: Circuit Detective Work ...................................................................... 111
Problem 3.25: Mystery Circuit ................................................................................... 111
Problem 3.26: More Circuit Detective Work ............................................................ 112
Problem 3.27: Linear, Time-Invariant Systems ....................................................... 113
Problem 3.28: Long and Sleepless Nights ............................................................... 114
Problem 3.29: A Testing Circuit ................................................................................. 114
Problem 3.30: Black-Box Circuit ................................................................................ 115
Problem 3.31: Solving a Mystery Circuit .................................................................. 115
Problem 3.32: Find the Load Impedance ................................................................. 116
Problem 3.33: Analog "Hum" Rejection ................................................................... 116
Problem3.34: An Interesting Circuit ......................................................................... 117
Problem 3.35: A Simple Circuit .................................................................................. 117
Problem 3.36: An Interesting and Useful Circuit ..................................................... 118
Problem 3.37: A Circuit Problem ............................................................................... 119
Problem 3.38: Analog Computers ............................................................................. 119
Problem 3.39: Transfer Functions and Circuits ....................................................... 120
Problem 3.40: Fun in the Lab .................................................................................... 120
Problem 3.41: Dependent Sources ........................................................................... 120
Problem 3.42: Operational Amplifers ....................................................................... 121
Problem 3.43: Op-Amp Circuit .................................................................................. 122
Problem 3.44: Why Op-Amps are Useful ................................................................. 123
Problem 3.45: Operational Amplifiers ...................................................................... 124
Problem 3.46: Designing a Bandpass Filter ............................................................. 124
Problem 3.47: Pre-emphasis or De-emphasis? ....................................................... 125
Problem 3.48: Active Filter ......................................................................................... 126
Problem 3.49: This is a filter? ..................................................................................... 126
Problem 3.50: Optical Receivers ............................................................................... 127
Problem 3.51: Reverse Engineering .......................................................................... 128
3.25 Solutions to Exercises in Chapter 3........................................................................... 128
Chapter 4 Frequency Domain ...................................................................................132
4.1 Introduction to the Frequency Domain ..................................................................... 132
4.2 Complex Fourier Series................................................................................................. 132
Exercise 4.2.1 ............................................................................................................... 133
Example 4.1 ................................................................................................................. 134
Exercise 4.2.2 ............................................................................................................... 137
4.3 Classic Fourier Series .................................................................................................... 138
Exercise 4.3.1 ............................................................................................................... 138
Exercise 4.3.2 ............................................................................................................... 140
Exercise 4.3.3 ............................................................................................................... 140
Example 4.2 ................................................................................................................. 141
4.4 A Signal's Spectrum ....................................................................................................... 142
Exercise 4.4.1 ............................................................................................................... 143
Exercise 4.4.2 ............................................................................................................... 144
4.5 Fourier Series Approximation of Signals .................................................................... 144
Exercise 4.5.1 ............................................................................................................... 148
4.6 Encoding Information in the Frequency Domain ...................................................... 149
Exercise 4.6.1 ............................................................................................................... 150
Exercise 4.6.2 ............................................................................................................... 151
4.7 Filtering Periodic Signals............................................................................................... 152
Example 4.3 ................................................................................................................. 153
Exercise 4.7.1 ............................................................................................................... 154
4.8 Derivation of the Fourier Transform ........................................................................... 154
Example 4.4 ................................................................................................................. 155
Exercise 4.8.1 ............................................................................................................... 156
Exercise 4.8.2 ............................................................................................................... 157
Example 4.5 ................................................................................................................. 159
Exercise 4.8.3 ............................................................................................................... 161
Exercise 4.8.4 ............................................................................................................... 161
4.9 Linear Time Invariant Systems..................................................................................... 161
Example 4.6 ................................................................................................................. 162
Exercise 4.9.1 ............................................................................................................... 163
4.9.1 Transfer Functions ............................................................................................. 164
4.9.2 Commutative Transfer Functions .................................................................... 164
4.10 Modeling the Speech Signal ....................................................................................... 165
Exercise 4.10.1 ............................................................................................................. 167
Exercise 4.10.2 ............................................................................................................. 169
Exercise 4.10.3 ............................................................................................................. 170
4.11 Frequency Domain Problems .................................................................................... 173
Problem 4.1: Simple Fourier Series .......................................................................... 173
Problem 4.2: Fourier Series ....................................................................................... 174
Problem 4.3: Phase Distortion .................................................................................. 175
Problem 4.4: Approximating Periodic Signals ......................................................... 176
Problem 4.5: Long, Hot Days ..................................................................................... 177
Problem 4.6: Fourier Transform Pairs ...................................................................... 178
Problem 4.7: Duality in Fourier Transforms ............................................................ 178
Problem 4.8: Spectra of Pulse Sequences ............................................................... 179
Problem 4.10: Lowpass Filtering a Square Wave .................................................... 180
Problem 4.11: Mathematics with Circuits ................................................................ 180
Problem 4.12: Where is that sound coming from? ................................................. 181
Problem 4.13: Arrangements of Systems ................................................................ 182
Problem 4.14: Filtering ............................................................................................... 182
Problem 4.15: Circuits Filter! ..................................................................................... 183
Problem 4.16: Reverberation .................................................................................... 183
Problem 4.17: Echoes in Telephone Systems .......................................................... 183
Problem 4.18: Effective Drug Delivery ...................................................................... 184
Problem 4.19: Catching Speeders with Radar ......................................................... 184
Problem 4.20: Demodulating an AM Signal ............................................................. 185
Problem 4.21: Unusual Amplitude Modulation ...................................................... 186
Problem 4.22: Sammy Falls Asleep... ........................................................................ 187
Problem 4.23: Jamming .............................................................................................. 188
Problem 4.24: AM Stereo ........................................................................................... 189
Problem 4.25: Novel AM Stereo Method ................................................................. 190
Problem 4.26: A Radical Radio Idea .......................................................................... 191
Problem 4.27: Secret Communication ..................................................................... 191
Problem 4.28: Signal Scrambling .............................................................................. 192
4.12 Solutions to Exercises in Chapter 4........................................................................... 192
Chapter 5 Digital Signal Processing .........................................................................196
5.1 Introduction to Digital Signal Processing.................................................................... 196
5.2 Introduction to Computer Organization .................................................................... 197
5.2.1 Computer Architecture ...................................................................................... 197
5.2.2 Representing Numbers ..................................................................................... 198
Exercise 5.2.1 ....................................................................................................... 199
Exercise 5.2.2 ....................................................................................................... 200
5.2.3 Computer Arithmetic and Logic ....................................................................... 200
Exercise 5.2.3 ....................................................................................................... 201
5.3 The Sampling Theorem ................................................................................................ 201
5.3.1 Analog-to-Digital Conversion ........................................................................... 201
5.4 The Sampling Theorem................................................................................................. 202
Exercise 5.3.1 ............................................................................................................... 204
Exercise 5.3.2 ............................................................................................................... 204
Exercise 5.3.3 ............................................................................................................... 205
5.5 Amplitude Quantization................................................................................................ 205
Exercise 5.4.1 ............................................................................................................... 206
Exercise 5.4.2 ............................................................................................................... 208
Exercise 5.4.3 ............................................................................................................... 208
Exercise 5.4.4 ............................................................................................................... 208
5.6 Discrete-Time Signals and Systems ............................................................................ 208
5.6.1 Real-and Complex-valued Signals .................................................................... 209
5.6.2 Complex Exponentials ....................................................................................... 209
5.6.3 Sinusoids ............................................................................................................ 209
5.6.4 Unit Sample......................................................................................................... 210
5.6.5 Unit Step ............................................................................................................. 210
5.6.6 Symbolic Signals ................................................................................................ 211
5.6.7 Discrete-Time Systems ..................................................................................... 211
5.7 Discrete-Time Fourier Transform (DTFT) .................................................................... 211
5.8 Discrete Fourier Transforms (DFT) .............................................................................. 217
5.9 DFT: Computational Complexity .................................................................................. 220
5.10 Fast Fourier Transform (FFT)...................................................................................... 221
5.11 Spectrograms............................................................................................................... 224
5.12 Discrete-Time Systems ............................................................................................... 228
5.13 Discrete-Time Systems in the Time-Domain ............................................................ 228
5.14 Discrete-Time Systems in the Frequency Domain .................................................. 233
5.14.1 Filtering in the Frequency Domain................................................................. 235
5.15 Efficiency of Frequency-Domain Filtering................................................................. 239
5.16 Discrete-Time Filtering of Analog Signals ................................................................. 242
5.17 Digital Signal Processing Problems ........................................................................... 244
5.18 Solutions to Exercises in Chapter 5........................................................................... 256
Chapter 6 Information Communication ..................................................................261
6.1 Information Communication ....................................................................................... 261
6.2 Types of Communication Channels ............................................................................ 262
6.3 Wireline Channels.......................................................................................................... 263
6.4 Wireless Channels ......................................................................................................... 269
6.5 Line-of-Sight Transmission ........................................................................................... 270
6.6 The Ionosphere and Communications ....................................................................... 272
6.7 Communication with Satellites .................................................................................... 272
6.8 Noise and Interference ................................................................................................. 273
6.9 Channel Models ............................................................................................................. 274
6.10 Baseband Communication......................................................................................... 276
6.11 Modulated Communication ....................................................................................... 277
6.12 Signal-to-Noise Ratio of an Amplitude-Modulated Signal ...................................... 278
6.13 Digital Communication ............................................................................................... 281
6.14 Binary Phase Shift Keying ........................................................................................... 282
6.15 Frequency Shift Keying ............................................................................................... 285
6.16 Digital Communication Receivers.............................................................................. 287
6.17 Digital Communication in the Presence of Noise.................................................... 289
6.18 Digital Communication System Properties .............................................................. 291
6.19 Digital Channels........................................................................................................... 292
6.20 Entropy ......................................................................................................................... 293
6.21 Source Coding Theorem ............................................................................................. 295
6.22 Compression and the Huffman Code ....................................................................... 297
6.23 Subtlies of Coding........................................................................................................ 299
6.24 Channel Coding ........................................................................................................... 302
6.25 Repetition Codes ......................................................................................................... 302
6.26 Block Channel Coding ................................................................................................. 304
6.27 Error-Correcting Codes: Hamming Distance............................................................ 305
6.28 Error-Correcting Codes: Channel Decoding ............................................................. 308
6.29 Error-Correcting Codes: Hamming Codes ................................................................ 310
6.30 Noisy Channel Coding Theorem ............................................................................... 313
6.30.1 Noisy Channel Coding Theorem .................................................................... 313
6.30.2 Converse to the Noisy Channel Coding Theorem ........................................ 314
6.31 Capacity of a Channel ................................................................................................. 315
6.32 Comparison of Analog and Digital Communication................................................ 316
6.33 Communication Networks ......................................................................................... 317
6.34 Message Routing ......................................................................................................... 319
6.35 Network architectures and interconnection ............................................................ 320
6.36 Ethernet ........................................................................................................................ 321
6.37 Communication Protocols.......................................................................................... 324
6.38 Information Communication Problems.................................................................... 325
6.39 Solutions to Exercises in Chapter 6........................................................................... 342
Chapter 7 Appendix ...................................................................................................351
7.1 Decibels .......................................................................................................................... 351
7.2 Permutations and Combinations ............................................................................... 352
7.2.1 Permutations and Combinations ..................................................................... 352
7.3 Frequency Allocations .................................................................................................. 354
7.4 Solutions to Exercises in Chapter 7 ............................................................................ 354
1
Chapter 1 Introduction
1.1 Themes
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
From its beginnings in the late nineteenth century, electrical engineering has
blossomed from focusing on electrical circuits for power, telegraphy and telephony to
focusing on a much broader range of disciplines. However, the underlying themes are
relevant today: Power creation and transmission and information have been the
underlying themes of electrical engineering for a century and a half. This course
concentrates on the latter theme: the representation, manipulation, transmission,
and reception of information by electrical means. This course describes what
information is, how engineers quantify information, and how electrical signals
represent information.
Information can take a variety of forms. When you speak to a friend, your thoughts
are translated by your brain into motor commands that cause various vocal tract
components the jaw, the tongue, the lips to move in a coordinated fashion.
Information arises in your thoughts and is represented by speech, which must have a
well defined, broadly known structure so that someone else can understand what you
say. Utterances convey information in sound pressure waves, which propagate to your
friend's ear. There, sound energy is converted back to neural activity, and, if what you
say makes sense, she understands what you say. Your words could have been
recorded on a compact disc (CD), mailed to your friend and listened to by her on her
stereo. Information can take the form of a text file you type into your word processor.
You might send the file via e-mail to a friend, who reads it and understands it. From an
information theoretic viewpoint, all of these scenarios are equivalent, although the
forms of the information representation sound waves, plastic and computer files are
very different.
Engineers, who don't care about information content, categorize information into two
different forms: analog and digital. Analog information is continuous valued;
examples are audio and video. Digital information is discrete valued; examples are
text (like what you are reading now) and DNA sequences.
The conversion of information-bearing signals from one energy form into another is
known as energyconversion or transduction. All conversion systems are inefficient
since some input energy is lost as heat, but this loss does not necessarily mean that
the conveyed information is lost. Conceptually we could use any form of energy to
represent information, but electric signals are uniquely well-suited for information
representation, transmission (signals can be broadcast from antennas or sent through
wires), and manipulation (circuits can be built to reduce noise and computers can be
used to modify information). Thus, we will be concerned with how to
• represent all forms of information with electrical signals,
• encode information as voltages, currents, and electromagnetic waves,
2
• manipulate information-bearing electric signals with circuits and computers, and
• receive electric signals and convert the information expressed by electric signals
back into a useful form.
Telegraphy represents the earliest electrical information system, and it dates from
1837. At that time, electrical science was largely empirical, and only those with
experience and intuition could develop telegraph systems. Electrical science came of
age when James Clerk Maxwell2 proclaimed in 1864 a set of equations that he claimed
governed all electrical phenomena. These equations predicted that light was an
electromagnetic wave, and that energy could propagate. Because of the complexity of
Maxwell's presentation, the development of the telephone in 1876 was due largely to
empirical work. Once Heinrich Hertz confirmed Maxwell's prediction of what we now
call radio waves in about 1882, Maxwell's equations were simplified by Oliver
Heaviside and others, and were widely read. This understanding of fundamentals led
to a quick succession of inventions the wireless telegraph (1899), the vacuum tube
(1905), and radio broadcasting that marked the true emergence of the
communications age.
During the first part of the twentieth century, circuit theory and electromagnetic
theory were all an electrical engineer needed to know to be qualified and produce
first-rate designs. Consequently, circuit theory served as the foundation and the
framework of all of electrical engineering education. At mid-century, three
"inventions" changed the ground rules. These were the first public demonstration of
the first electronic computer (1946), the invention of the transistor (1947), and the
publication of A Mathematical Theoryof Communication by Claude Shannon (1948).
Although conceived separately, these creations gave birth to the information age, in
which digital and analog communication systems interact and compete for design
preferences. About twenty years later, the laser was invented, which opened even
more design possibilities. Thus, the primary focus shifted from how to build
communication systems (the circuit theory era) to what communications systems
were intended to accomplish. Only once the intended system is specified can an
implementation be selected. Today's electrical engineer must be mindful of the
system's ultimate goal, and understand the tradeoffs between digital and analog
alternatives, and between hardware and software configurations in designing
information systems.
Note: Thanks to the translation efforts of Rice University's Disability Support
Services , this collection is now available in a Braille-printable version. Please
click here5 to download a .zip file containing all the necessary .dxb and image
files.
1.2 Signals Represent Information
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Whether analog or digital, information is represented by the fundamental quantity in
electrical engineering: the signal. Stated in mathematical terms, a signal is merely a
function. Analog signals are continuous-valued; digital signals are discrete-valued.
3
The independent variable of the signal could be time (speech, for example), space
(images), or the integers (denoting the sequencing of letters and numbers in the
football score).
1.2.1 Analog Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Analog signals are usually signals defined over continuous independent
variable(s). Modeling the Speech Signal (Page 165) is produced by your vocal cords
exciting acoustic resonances in your vocal tract. The result is pressure waves
propagating in the air, and the speech signal thus corresponds to a function having
independent variables of space and time and a value corresponding to air pressure: s
(x, t) (Here we use vector notation x to denote spatial coordinates). When you record
someone talking, you are evaluating the speech signal at a particular spatial location,
x0 say. An example of the resulting waveform s (x0,t) is shown in this Figure 1.1.
Speech Example
Figure 1.1 Speech Example A speech signal's amplitude relates to tiny air pressure variations. Shown is a
recording of the vowel "e" (as in "speech").
Photographs are static, and are continuous-valued signals defined over space. Blackand-white images have only one value at each point in space, which amounts to its
optical refection properties. In Figure 1.2, an image is shown, demonstrating that it
(and all other images as well) are functions of two independent spatial variables.
4
Figure 1.2 Lena On the left is the classic Lena image, which is used ubiquitously as a test image. It contains
straight and curved lines, complicated texture, and a face. On the right is a perspective display of the Lena
image as a signal: a function of two spatial variables. The colors merely help show what signal values are
about the same size. In this image, signal values range between 0 and 255; why is that?
Color images have values that express how reflectivity depends on the optical
spectrum. Painters long ago found that mixing together combinations of the so-called
primary colors red, yellow and blue can produce very realistic color images. Thus,
images today are usually thought of as having three values at every point in space, but
a different set of colors is used: How much of red, green and blue is present.
Mathematically, color pictures are multivalued vector-valued signals: s (x)=(r (x) ,g (x) ,b
T
(x)) .
Interesting cases abound where the analog signal depends not on a continuous
variable, such as time, but on a discrete variable. For example, temperature readings
taken every hour have continuous analog values, but the signal's independent variable
is (essentially) the integers.
1.2.2 Digital Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The word "digital" means discrete-valued and implies the signal has an integer-valued
independent variable. Digital information includes numbers and symbols (characters
typed on the keyboard, for example). Computers rely on the digital representation of
information to manipulate and transform information. Symbols do not have a
numeric value, and each is represented by a unique number. The ASCII character code
has the upper-and lowercase characters, the numbers, punctuation marks, and
various other symbols represented by a seven-bit integer. For example, the ASCII code
represents the letter a as the number 97 and the letter A as 65. Figure 1.3 shows the
international convention on associating characters with integers.
5
Figure 1.3 ASCII Table The ASCII translation table shows how standard keyboard characters are
represented by integers. In pairs of columns, this table displays first the so-called 7-bit code (how many
characters in a seven-bit code?), then the character the number represents. The numeric codes are
represented in hexadecimal (base-16) notation. Mnemonic characters correspond to control characters,
some of which may be familiar (like cr for carriage return) and some not (bel means a "bell").
1.3 Structure of Communication Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 1.4 The Fundamental Model of Communication.
Figure 1.5 Definition of a system A system operates on its input signal x (t) to produce an output y (t).
The fundamental model of communications is portrayed in Figure 1.4. In this
fundamental model, each message-bearing signal, exemplified by s (t), is analog and is
a function of time. A system operates on zero, one, or several signals to produce
more signals or to simply absorb them Figure 1.5. In electrical engineering, we
represent a system as a box, receiving input signals (usually coming from the left) and
producing from them new output signals. This graphical representation is known as a
block diagram. We denote input signals by lines having arrows pointing into the box,
output signals by arrows pointing away. As typified by the communications model,
6
how information flows, how it is corrupted and manipulated, and how it is ultimately
received is summarized by interconnecting block diagrams: The outputs of one or
more systems serve as the inputs to others.
In the communications model, the source produces a signal that will be absorbed by
the sink. Examples of time-domain signals produced by a source are music, speech,
and characters typed on a keyboard. Signals can also be functions of two variables an
image is a signal that depends on two spatial variables or more television pictures
(video signals) are functions of two spatial variables and time. Thus, information
sources produce signals. In physical systems, each signal corresponds to an
electrical voltage or current. To be able to design systems, we must understand
electrical science and technology. However, we first need to understand the big
picture to appreciate the context in which the electrical engineer works.
In communication systems, messages signals produced by sourcesmust be recast for
transmission. The block diagram has the message s (t) passing through a block
labeled transmitter that produces the signal x (t). In the case of a radio transmitter, it
accepts an input audio signal and produces a signal that physically is an
electromagnetic wave radiated by an antenna and propagating as Maxwell's equations
predict. In the case of a computer network, typed characters are encapsulated in
packets, attached with a destination address, and launched into the Internet. From the
communication systems "big picture" perspective, the same block diagram applies
although the systems can be very different. In any case, the transmitter should not
operate in such a way that the message s (t) cannot be recovered from x (t). In the
mathematical sense, the inverse system must exist, else the communication system
cannot be considered reliable. (It is ridiculous to transmit a signal in such a way that
no one can recover the original. However, clever systems exist that transmit signals so
that only the "in crowd" can recover them. Such crytographic systems underlie secret
communications.)
Transmitted signals next pass through the next stage, the evil channel. Nothing good
happens to a signal in a channel: It can become corrupted by noise, distorted, and
attenuated among many possibilities. The channel cannot be escaped (the real world
is cruel), and transmitter design and receiver design focus on how best to jointly fend
of the channel's effects on signals. The channel is another system in our block
diagram, and produces r (t), the signal received by the receiver. If the channel were
benign (good luck finding such a channel in the real world), the receiver would serve
as the inverse system to the transmitter, and yield the message with no distortion.
However, because of the channel, the receiver must do its best to produce a received
message sˆ(t) that resembles s (t) as much as possible. Shannon8 showed in his 1948
paper that reliable for the moment, take this word to mean error-free digital
communication was possible over arbitrarily noisy channels. It is this result that
modern communications systems exploit, and why many communications systems
are going "digital." The module on Information Communication (Section 6.1) details
Shannon's theory of information, and there we learn of Shannon's result and how to
use it.
Finally, the received message is passed to the information sink that somehow makes
use of the message. In the communications model, the source is a system having no
input but producing an output; a sink has an input and no output.
7
Understanding signal generation and how systems work amounts to understanding
signals, the nature of the information they represent, how information is transformed
between analog and digital forms, and how information can be processed by systems
operating on information-bearing signals. This understanding demands two different
fields of knowledge. One is electrical science: How are signals represented and
manipulated electrically? The second is signal science: What is the structure of signals,
no matter what their source, what is their information content, and what capabilities
does this structure force upon communication systems?
1.4 The Fundamental Signal
1.4.1 The Sinusoid
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The most ubiquitous and important signal in electrical engineering is the sinusoid.
Sine Definition
A is known as the sinusoid's amplitude, and determines the sinusoid's size. The
amplitude conveys the sinusoid's physical units (volts, lumens, etc). The frequencyf
-1
has units of Hz (Hertz) or s , and determines how rapidly the sinusoid oscillates per
unit time. The temporal variable t always has units of seconds, and thus the frequency
determines how many oscillations/second the sinusoid has. AM radio stations have
carrier frequencies of about 1 MHz (one mega-hertz or 106 Hz), while FM stations have
carrier frequencies of about 100 MHz. Frequency can also be expressed by the symbol
ω, which has units of radians/second. Clearly,
. In communications, we most often express frequency in Hertz. Finally, φ is the
phase, and determines the sine wave's behavior at the origin (t =0). It has units of
radians, but we can express it in degrees, realizing that in computations we must
convert from degrees to radians. Note that if
, the sinusoid corresponds to a sine function, having a zero value at the origin.
Thus, the only difference between a sine and cosine signal is the phase; we term
either a sinusoid.
We can also define a discrete-time variant of the sinusoid:
. Here, the independent variable is n and represents the integers. Frequency now has
no dimensions, and takes on values between 0 and 1.
8
Exercise 1.4.1
Show that cos (2πfn) = cos (2π (f + 1) n), which means that a sinusoid
having a frequency larger than one corresponds to a sinusoid having
a frequency less than one.
Note: Notice that we shall call either sinusoid an analog signal. Only when the
discrete-time signal takes on a finite set of values can it be considered a digital
signal.
Exercise 1.4.2
Can you think of a simple signal that has a finite number of values
but is defined in continuous time? Such a signal is also an analog
signal.
1.4.2 Communicating Information with Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The basic idea of communication engineering is to use a signal's parameters to
represent either real numbers or other signals. The technical term is to modulate the
carrier signal's parameters to transmit information from one place to another. To
explore the notion of modulation, we can send a real number (today's temperature,
for example) by changing a sinusoid's amplitude accordingly. If we wanted to send the
daily temperature, we would keep the frequency constant (so the receiver would know
what to expect) and change the amplitude at midnight. We could relate temperature
to amplitude by the formula A = A0(1 + kT), where A0 and k are constants that the
transmitter and receiver must both know.
If we had two numbers we wanted to send at the same time, we could modulate the
sinusoid's frequency as well as its amplitude. This modulation scheme assumes we
can estimate the sinusoid's amplitude and frequency; we shall learn that this is indeed
possible.
Now suppose we have a sequence of parameters to send. We have exploited all of the
sinusoid's two parameters. What we can do is modulate them for a limited time (say T
seconds), and send two parameters every T. This simple notion corresponds to how a
modem works. Here, typed characters are encoded into eight bits, and the individual
bits are encoded into a sinusoid's amplitude and frequency. We'll learn how this is
done in subsequent modules, and more importantly, we'll learn what the limits are on
such digital communication schemes.
9
1.5 Introduction Problems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Problem 1.1: RMS Values
The rms (root-mean-square) value of a periodic signal is defined to be
where T is defined to be the signal's period: the smallest positive number such that s
(t)= s (t + T ).
1. What is the period of s (t)= Asin (2πf0t + φ)?
2. What is the rms value of this signal? How is it related to the peak value?
3. What is the period and rms value of the depicted (Figure 1.6) square wave,
generically denoted by sq (t)?
4. By inspecting any device you plug into a wall socket, you'll see that it is labeled
"110 volts AC". What is the expression for the voltage provided by a wall socket?
What is its rms value?
Figure 1.6
Problem 1.2: Modems
The word "modem" is short for "modulator-demodulator." Modems are used not only
for connecting computers to telephone lines, but also for connecting digital (discretevalued) sources to generic channels. In this problem, we explore a simple kind of
modem, in which binary information is represented by the presence or absence of a
sinusoid (presence representing a "1" and absence a "0"). Consequently, the modem's
transmitted signal that represents a single bit has the form
Within each bit interval T, the amplitude is either A or zero.
1. What is the smallest transmission interval that makes sense with the frequency
f 0?
2. Assuming that ten cycles of the sinusoid comprise a single bit's transmission
interval, what is the datarate of this transmission scheme?
10
3. Now suppose instead of using "on-of" signaling, we allow one of several different
values for the amplitude during any transmission interval. If N amplitude values
are used, what is the resulting datarate?
4. The classic communications block diagram applies to the modem. Discuss how
the transmitter must interface with the message source since the source is
producing letters of the alphabet, not bits.
Problem 1.3: Advanced Modems
To transmit symbols, such as letters of the alphabet, RU computer modems use two
frequencies (1600 and 1800 Hz) and several amplitude levels. A transmission is sent
for a period of time T (known as the transmission or baud interval) and equals the
sum of two amplitude-weighted carriers.
We send successive symbols by choosing an appropriate frequency and amplitude
combination, and sending them one after another.
1. What is the smallest transmission interval that makes sense to use with the
frequencies given above? In other words, what should T be so that an integer
number of cycles of the carrier occurs?
2. Sketch (using Matlab) the signal that modem produces over several transmission
intervals. Make sure you axes are labeled.
3. Using your signal transmission interval, how many amplitude levels are needed to
transmit ASCII characters at a datarate of 3,200 bits/s? Assume use of the
extended (8-bit) ASCII code.
Note: We use a discrete set of values for A1 and A2. If we have N1 values for
amplitude A1, and N2 values for A2, we have N1N2 possible symbols that can be
sent during each T second interval. Toconvert this number into bits (the
fundamental unit of information engineers use to qualify things), compute
log2 (N1N2).
1.6 Solutions to Exercises in Chapter 1
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 1.4.1
Solution to Exercise 1.4.2
A square wave takes on the values 1 and −1 alternately. See the plot in the module
Square Wave (Page 20)
11
Chapter 2 Signals and Systems
2.1 Complex Numbers
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
While the fundamental signal used in electrical engineering is the sinusoid, it can be
expressed mathematically in terms of an even more fundamental signal: the complex
exponential. Representing sinusoids in terms of complex exponentials is not a
mathematical oddity. Fluency with complex numbers and rational functions of
complex variables is a critical skill all engineers master. Understanding information
and power system designs and developing new systems all hinge on using complex
numbers. In short, they are critical to modern electrical engineering, a realization
made over a century ago.
2.1.1 Definitions
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The notion of the square root of −1 originated with the quadratic formula: the solution
of certain quadratic equations mathematically exists only if the so-called imaginary
quantity
could be defined. Euler first used i for the imaginary unit but that notation did not take
hold until roughly Ampere's time. Ampere3 used the symbol i to denote current
(intensite de current). It wasn't until the twentieth century that the importance of
complex numbers to circuit theory became evident. By then, using i for current was
entrenched and electrical engineers chose j for writing complex numbers.
An imaginary number has the form
.A complex number, z, consists of the ordered pair (a,b), a is the real component and
b is the imaginary component (the j is suppressed because the imaginary component
of the pair is always in the second position). The imaginary number jb equals (0,b).
Note that a and b are real-valued numbers.
Figure 2.1 shows that we can locate a complex number in what we call the complex
plane. Here, a, the real part, is the x-coordinate and b, the imaginary part, is the ycoordinate.
From analytic geometry, we know that locations in the plane can be expressed as the
sum of vectors, with the vectors corresponding to the x and y directions.
Consequently, a complex number z can be expressed as the (vector) sum z = a + jb
where j indicates the y-coordinate. This representation is known as the Cartesian
form of z. An imaginary number can't be numerically added to a real number; rather,
12
this notation for a complex number represents vector addition, but it provides a
convenient notation when we perform arithmetic manipulations.
Some obvious terminology. The real part of the complex number z = a + jb, written as
Re (z), equals a. We consider the real part as a function that works by selecting that
component of a complex number not multiplied by j. The imaginary part of z, Im (z),
equals b: that part of a complex number that is multiplied by j. Again, both the real
and imaginary parts of a complex number are real-valued.
Figure 2.1 The Complex Plane A complex number is an ordered pair (a,b) that can be regarded as
coordinates in the plane. Complex numbers can also be expressed in polar coordinates as r∠θ.
The complex conjugate of z, written as z *, has the same real part as z but an
imaginary part of the opposite sign.
Using Cartesian notation, the following properties easily follow.
• If we add two complex numbers, the real part of the result equals the sum of the
real parts and the imaginary part equals the sum of the imaginary parts. This
property follows from the laws of vector addition.
In this way, the real and imaginary parts remain separate.
• The product of j and a real number is an imaginary number: ja. The product of j
2
and an imaginary number is a real number: j (jb)= −b because j = −1.
13
Consequently, multiplying a complex number by j rotates the number's position
by 90 degrees.
Exercise 2.1.1
Use the Definition of addition to show that the real and imaginary
parts can be expressed as a sum/diference of a complex number and
its conjugate.
and
Complex numbers can also be expressed in an alternate form, polar form, which we
will find quite useful. Polar form arises arises from the geometric interpretation of
complex numbers. The Cartesian form of a complex number can be re-written as
By forming a right triangle having sides a and b, we see that the real and imaginary
parts correspond to the cosine and sine of the triangle's base angle. We thus obtain
the polar form for complex numbers.
The quantity r is known as the magnitude of the complex number z, and is frequently
written as |z|. The quantity θ is the complex number's angle. In using the arc-tangent
formula to find the angle, we must take into account the quadrant in which the
complex number lies.
Exercise 2.1.2
Convert 3 − 2j to polar form.
2.1.2 Euler's Formula
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Surprisingly, the polar form of a complex number z can be expressed mathematically
as
14
To show this result, we use Euler's relations that express exponentials with imaginary
arguments in terms of trigonometric functions.
The first of these is easily derived from the Taylor's series for the exponential.
Substituting jθ for x, we find that
2
3
4
because j = −1, j = −j, and j =1. Grouping separately the real-valued terms and the
imaginary-valued ones,
The real-valued terms correspond to the Taylor's series for cos (θ), the imaginary ones
to sin (θ), and Euler's first relation results. The remaining relations are easily derived
from the first. We see that multiplying the exponential in (2.3) by a real constant
corresponds to setting the radius of the complex number to the constant.
2.1.3 Calculating with Complex Numbers
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Adding and subtracting complex numbers expressed in Cartesian form is quite easy:
You add (subtract) the real parts and imaginary parts separately.
To multiply two complex numbers in Cartesian form is not quite as easy, but follows
directly from following the usual rules of arithmetic.
Note that we are, in a sense, multiplying two vectors to obtain another vector.
Complex arithmetic provides a unique way of defining vector multiplication.
15
Exercise 2.1.3
What is the product of a complex number and its conjugate?
Division requires mathematical manipulation. We convert the
division problem into a multiplication problem by multiplying both
the numerator and denominator by the conjugate of the
denominator.
Because the final result is so complicated, it's best to remember how to perform
division multiplying numerator and denominator by the complex conjugate of the
denominator than trying to remember the final result.
The properties of the exponential make calculating the product and ratio of two
complex numbers much simpler when the numbers are expressed in polar form.
To multiply, the radius equals the product of the radii and the angle the sum of the
angles. To divide, the radius equals the ratio of the radii and the angle the difference
of the angles. When the original complex numbers are in Cartesian form, it's usually
worth translating into polar form, then performing the multiplication or division
(especially in the case of the latter). Addition and subtraction of polar forms amounts
to converting to Cartesian form, performing the arithmetic operation, and converting
back to polar form.
16
Example 2.1
When we solve circuit problems, the crucial quantity, known as a
transfer function, will always be expressed as the ratio of
polynomials in the variable s = j2πf. What we'll need to understand
the circuit's effect is the transfer function in polar form. For
instance, suppose the transfer function equals
Performing the required division is most easily accomplished by first
expressing the numerator and denominator each in polar form, then
calculating the ratio. Thus,
2.2 Elemental Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Elemental signals are the building blocks with which we build complicated
signals. By definition, elemental signals have a simple structure. Exactly what we
mean by the "structure of a signal" will unfold in this section of the course. Signals are
nothing more than functions defined with respect to some independent variable,
which we take to be time for the most part. Very interesting signals are not functions
solely of time; one great example of which is an image. For it, the independent
variables are x and y (two-dimensional space). Video signals are functions of three
variables: two spatial dimensions and time. Fortunately, most of the ideas underlying
modern signal theory can be exemplified with one-dimensional signals.
2.2.1 Sinusoids
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Perhaps the most common real-valued signal is the sinusoid.
For this signal, A is its amplitude, f0 its frequency, and φ its phase.
17
2.2.2 Complex Exponentials
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The most important signal is complex-valued, the complex exponential.
Here, j denotes
is known as the signal's complex amplitude. Considering the complex amplitude as a
complex number in polar form, its magnitude is the amplitude A and its angle the
signal phase. The complex amplitude is also known as a phasor. The complex
exponential cannot be further decomposed into more elemental signals, and is the
most important signal in electrical engineering! Mathematical manipulations at
first appear to be more difficult because complex-valued numbers are introduced. In
fact, early in the twentieth century, mathematicians thought engineers would not be
sufficiently sophisticated to handle complex exponentials even though they greatly
simplified solving circuit problems. Steinmetz 5 introduced complex exponentials to
electrical engineering, and demonstrated that "mere" engineers could use them to
good effect and even obtain right answers! See Complex Numbers (Page 11) for a
review of complex numbers and complex arithmetic.
The complex exponential defines the notion of frequency: it is the only signal that
contains only one frequency component. The sinusoid consists of two frequency
components: one at the frequency +f0 and the other at −f0.
EULER RELATION: This decomposition of the sinusoid can be traced to Euler's relation.
DECOMPOSITION: The complex exponential signal can thus be written in terms of its
real and imaginary parts using Euler's relation. Thus, sinusoidal signals can be
expressed as either the real or the imaginary part of a complex exponential signal, the
choice depending on whether cosine or sine phase is needed, or as the sum of two
complex exponentials. These two decompositions are mathematically equivalent to
each other.
18
Using the complex plane, we can envision the complex exponential's temporal
variations as seen in the above Figure 2.2. The magnitude of the complex exponential
is A, and the initial value of the complex exponential at t =0 has an angle of φ. As time
increases, the locus of points traced by the complex exponential is a circle (it has
constant magnitude of A). The number of times per second we go around the circle
equals the frequency f. The time taken for the complex exponential to go around the
circle once is known as its period T, and equals
. The projections onto the real and imaginary axes of the rotating vector representing
the complex exponential signal are the cosine and sine signal of Euler's relation
((2.16)).
2.2.3 Real Exponentials
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
As opposed to complex exponentials which oscillate, real exponentials (Figure 2.3)
decay.
The quantity τ is known as the exponential's time constant, and corresponds to the
time required for the exponential to decrease by a factor of
, which approximately equals 0.368. A decaying complex exponential is the product
of a real and a complex exponential.
In the complex plane, this signal corresponds to an exponential spiral. For such
signals, we can define complex frequency as the quantity multiplying t.
2.2.4 Unit Step
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The unit step function (Figure 2.4) is denoted by u (t), and is defined to be
19
Figure 2.2 The complex exponential Graphically, the complex exponential scribes a circle in the complex
plane as time evolves. Its real and imaginary parts are sinusoids. The rate at which the signal goes around
the circle is the frequency f and the time taken to go around is the period T . A fundamental relationship is T
=f1.
Figure 2.3 The real exponential.
Figure 2.4 The unit step.
20
Caution: This signal is discontinuous at the origin. Its value at the origin
need not be defined, and doesn't matter in signal theory.
This kind of signal is used to describe signals that "turn on" suddenly. For example, to
mathematically represent turning on an oscillator, we can write it as the product of a
sinusoid and a step: s (t)= Asin (2πft) u (t).
2.2.5 Pulse
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 2.5 describes turning a unit-amplitude signal on for a duration of Δ seconds,
then turning it off.
Figure 2.5 The pulse.
We will find that this is the second most important signal in communications.
2.2.6 Square Wave
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 2.6 sq (t) is a periodic signal like the sinusoid. It too has an amplitude and a
period, which must be specifed to characterize the signal. We find subsequently that
the sine wave is a simpler signal than the square wave.
Figure 2.6 The square wave
21
2.2.7 Signal Decomposition
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A signal's complexity is not related to how wiggly it is. Rather, a signal expert looks for
ways of decomposing a given signal into a sum of simpler signals, which we term the
signal decomposition. Though we will never compute a signal's complexity, it
essentially equals the number of terms in its decomposition. In writing a signal as a
sum of component signals, we can change the component signal's gain by multiplying
it by a constant and by delaying it. More complicated decompositions could contain
derivatives or integrals of simple signals.
Example 2.2
As an example of signal complexity, we can express the pulse pΔ (t) as
a sum of delayed unit steps.
Thus, the pulse is a more complex signal than the step. Be that as it
may, the pulse is very useful to us.
Exercise 2.3.1
Express a square wave having period T and amplitude A as a
superposition of delayed and amplitude-scaled pulses.
Because the sinusoid is a superposition of two complex exponentials, the sinusoid is
more complex. We could not prevent ourselves from the pun in this statement.
Clearly, the word "complex" is used in two different ways here. The complex
exponential can also be written (using Euler's relation (2.16)) as a sum of a sine and a
cosine. We will discover that virtually every signal can be decomposed into a sum of
complex exponentials, and that this decomposition is very useful. Thus, the complex
exponential is more fundamental, and Euler's relation does not adequately reveal its
complexity.
2.3 Discrete-Time Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
So far, we have treated what are known as analog signals and systems.
Mathematically, analog signals are functions having continuous quantities as their
independent variables, such as space and time. Discrete-time signals (Section 5.5) are
functions defined on the integers; they are sequences. One of the fundamental results
of signal theory (Section 5.3) will detail conditions under which an analog signal can be
converted into a discrete-time one and retrieved without error. This result is
22
important because discrete-time signals can be manipulated by systems instantiated
as computer programs. Subsequent modules describe how virtually all analog signal
processing can be performed with software.
As important as such results are, discrete-time signals are more general,
encompassing signals derived from analog ones and signals that aren't. For example,
the characters forming a text file form a sequence, which is also a discrete-time signal.
We must deal with such symbolic valued (p. 180) signals and systems as well.
As with analog signals, we seek ways of decomposing real-valued discrete-time signals
into simpler components. With this approach leading to a better understanding of
signal structure, we can exploit that structure to represent information (create ways of
representing information with signals) and to extract information (retrieve the
information thus represented). For symbolic-valued signals, the approach is different:
We develop a common representation of all symbolic-valued signals so that we can
embody the information they contain in a unifed way. From an information
representation perspective, the most important issue becomes, for both real-valued
and symbolic-valued signals, efciency; What is the most parsimonious and compact
way to represent information so that it can be extracted later.
2.3.1 Real-and Complex-valued Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A discrete-time signal is represented symbolically as s (n), where n = {..., −1, 0, 1,... }.
We usually draw discrete-time signals as stem plots to emphasize the fact they are
functions defined only on the integers. We can delay a discrete-time signal by an
integer just as with analog ones. A delayed unit sample has the expression δ (n − m),
and equals one when n = m.
Figure 2.7 Discrete-Time Cosine Signal The discrete-time cosine signal is plotted as a stem plot. Can you
find the formula for this signal?
2.3.2 Complex Exponentials
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The most important signal is, of course, the complex exponential sequence.
23
2.3.3 Sinusoids
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Discrete-time sinusoids have the obvious form
. As opposed to analog complex exponentials and sinusoids that can have their
frequencies be any real value, frequencies of their discrete time counterparts yield
unique waveforms only when f lies in the interval
. This property can be easily understood by noting that adding an integer to the
frequency of the discrete-time complex exponential has no effect on the signal's value.
This derivation follows because the complex exponential evaluated at an integer
multiple of 2π equals one.
2.3.4 Unit Sample
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The second-most important discrete-time signal is the unit sample, which is defined
to be
Unit Sample
Figure 2.8 Unit Sample
Examination of a discrete-time signal's plot, like that of the cosine signal shown in
Figure 2.7 (Discrete-Time Cosine Signal), reveals that all signals consist of a sequence
of delayed and scaled unit samples. Because the value of a sequence at each integer
m is denoted by s (m) and the unit sample delayed to occur at m is written δ (n − m),
we can decompose any signal as a sum of unit samples delayed to the appropriate
location and scaled by the signal value.
24
This kind of decomposition is unique to discrete-time signals, and will prove useful
subsequently.
Discrete-time systems can act on discrete-time signals in ways similar to those found
in analog signals and systems. Because of the role of software in discrete-time
systems, many more different systems can be envisioned and "constructed" with
programs than can be with analog signals. In fact, a special class of analog signals can
be converted into discrete-time signals, processed with software, and converted back
into an analog signal, all without the incursion of error. For such signals, systems can
be easily produced in software, with equivalent analog realizations difficult, if not
impossible, to design.
2.3.5 Symbolic-valued Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Another interesting aspect of discrete-time signals is that their values do not need to
be real numbers. We do have real-valued discrete-time signals like the sinusoid, but
we also have signals that denote the sequence of characters typed on the keyboard.
Such characters certainly aren't real numbers, and as a collection of possible signal
values, they have little mathematical structure other than that they are members of a
set. More formally, each element of the symbolic-valued signal s (n) takes on one of
the values {a1,...,aK} which comprise the alphabet A. This technical terminology does
not mean we restrict symbols to being members of the English or Greek alphabet.
They could represent keyboard characters, bytes (8-bit quantities), integers that
convey daily temperature. Whether controlled by software or not, discrete-time
systems are ultimately constructed from digital circuits, which consist entirely of
analog circuit elements. Furthermore, the transmission and reception of discrete-time
signals, like e-mail, is accomplished with analog signals and systems. Understanding
how discrete-time and analog signals and systems intertwine is perhaps the main goal
of this course.
 MEDIA OBJECT (http://www.opentextbooks.org.hk/system/files/resource/9/9648/9678/media/
discrete-time-signals-25.pdf)
2.3.6 Introduction to Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Signals are manipulated by systems. Mathematically, we represent what a system
does by the notation y (t)= S (x (t)), with x representing the input signal and y the output
signal.
25
system.
Figure 2.9 Definition of a system The system depicted has input x (t) and output y (t). Mathematically,
systems operate on function(s) to produce other function(s). In many ways, systems are like functions, rules
that yield a value for the dependent variable (our output signal) for each value of its independent variable
(its input signal). The notation y (t)= S (x (t)) corresponds to this block diagram. We term S (·) the input-output
relation for the
This notation mimics the mathematical symbology of a function: A system's input is
analogous to an independent variable and its output the dependent variable. For the
mathematically inclined, a system is a functional: a function of a function (signals are
functions).
Simple systems can be connected together one system's output becomes another's
input to accomplish some overall design. Interconnection topologies can be quite
complicated, but usually consist of weaves of three basic interconnection forms.
2.3.6.1 Cascade Interconnection
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Cascade
Figure 2.10 Fundamental model of communication The most rudimentary ways of interconnecting
systems are shown in the fgures in this section. This is the cascade confguration.
The simplest form is when one system's output is connected only to another's input.
Mathematically, w (t)= S1 (x (t)), and y (t)= S2 (w (t)), with the information contained in x
(t) processed by the first, then the second system. In some cases, the ordering of the
systems matter, in others it does not. For example, in the fundamental model of
communication (Figure 2.10) the ordering most certainly matters.
2.3.7 Parallel Interconnection
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
parallel
26
Figure 2.11 The parallel confguration.
A signal x (t) is routed to two (or more) systems, with this signal appearing as the input
to all systems simultaneously and with equal strength. Block diagrams have the
convention that signals going to more than one system are not split into pieces along
the way. Two or more systems operate on x (t) and their outputs are added together
to create the output y (t). Thus, y (t)= S1 (x (t))+S2 (x (t)), and the information in x (t) is
processed separately by both systems.
2.3.8 Feedback Interconnection
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
feedback
Figure 2.12 The feedback configuration.
The subtlest interconnection configuration has a system's output also contributing to
its input. Engineers would say the output is "fed back" to the input through system 2,
hence the terminology. The mathematical statement of the feedback interconnection
(Figure 2.12) is that the feed-forward system produces the output: y (t)= S1 (e (t)). The
input e (t) equals the input signal minus the output of some other system's output to y
(t): e (t)= x (t) − S2 (y (t)). Feedback systems are omnipresent in control problems, with
the error signal used to adjust the output to achieve some condition defined by the
input (controlling) signal.
For example, in a car's cruise control system, x (t) is a constant representing what
speed you want, and y (t) is the car's speed as measured by a speedometer. In this
application, system 2 is the identity system (output equals input).
27
2.4 Simple Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Systems manipulate signals, creating output signals derived from their inputs. Why
the following are categorized as "simple" will only become evident towards the end of
the course.
2.4.1 Sources
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Sources produce signals without having input. We like to think of these as having
controllable parameters, like amplitude and frequency. Examples would be oscillators
that produce periodic signals like sinusoids and square waves and noise generators
that yield signals with erratic waveforms (more about noise subsequently). Simply
writing an expression for the signals they produce specifies sources. A sine wave
generator might be specified by y (t)= Asin (2πf0t) u (t), which says that the source
was turned on at t =0 to produce a sinusoid of amplitude A and frequency f0.
2.4.2 Amplifiers
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
An amplifier (Figure 2.13) multiplies its input by a constant known as the amplifier
gain.
Figure 2.13 An amplifier.
The gain can be positive or negative (if negative, we would say that the amplifier
inverts its input) and its magnitude can be greater than one or less than one. If less
than one, the amplifier actually attenuates. A real-world example of an amplifier is
your home stereo. You control the gain by turning the volume control.
28
2.4.3 Delay
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A system serves as a time delay (Figure 2.14) when the output signal equals the input
signal at an earlier time.
Figure 2.14 A delay.
Here, τ is the delay. The way to understand this system is to focus on the time origin:
The output at time t = τ equals the input at time t =0. Thus, if the delay is positive, the
output emerges later than the input, and plotting the output amounts to shifting the
input plot to the right. The delay can be negative, in which case we say the system
advances its input. Such systems are difficult to build (they would have to produce
signal values derived from what the input will be), but we will have occasion to
advance signals in time.
2.4.4 Time Reversal
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Here, the output signal equals the input signal flipped about the time origin.
Figure 2.15 A time reversal system.
Again, such systems are difficult to build, but the notion of time reversal occurs
frequently in communications systems.
Exercise 2.6.1
Mentioned earlier was the issue of whether the ordering of systems
mattered. In other words, if we have two systems in cascade, does the
output depend on which comes first? Determine if the ordering
matters for the cascade of an amplifier and a delay and for the
cascade of a time-reversal system and a delay.
29
2.4.5 Derivative Systems and Integrators
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Systems that perform calculus-like operations on their inputs can produce waveforms
signifcantly different than present in the input. Derivative systems operate in a
straightforward way: A first-derivative system would have the input-output
relationship
. Integral systems have the complication that the integral's limits must be defined. It is
a signal theory convention that the elementary integral operation have a lower limit of
−∞, and that the value of all signals at t = −∞ equals zero. A simple integrator would
have input-output relation
2.4.6 Linear Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Linear systems are a class of systems rather than having a specific input-output
relation. Linear systems form the foundation of system theory, and are the most
important class of systems in communications. They have the property that when the
input is expressed as a weighted sum of component signals, the output equals the
same weighted sum of the outputs produced by each component. When S (·) is linear,
for all choices of signals and gains. This general input-output relation property can be
manipulated to indicate specific properties shared by all linear systems.
• S (Gx (t)) = GS (x (t)) The colloquialism summarizing this property is "Double the
input, you double the output." Note that this property is consistent with alternate
ways of expressing gain changes: Since 2x (t) also equals x (t)+ x (t), the linear
system Definition provides the same output no matter which of these is used to
express a given signal.
• S (0) =0 If the input is identically zero for all time, the output of a linear system
must be zero. This property follows from the simple derivation S (0) = S (x (t) − x
(t)) = S (x (t)) − S (x (t)) = 0.
Just why linear systems are so important is related not only to their properties, which
are divulged throughout this course, but also because they lend themselves to
relatively simple mathematical analysis. Said another way, "They're the only systems
we thoroughly understand!"
We can find the output of any linear system to a complicated input by decomposing
the input into simple signals. The equation above (2.34) says that when a system is
30
linear, its output to a decomposed input is the sum of outputs to each input. For
example, if
the output S (x (t)) of any linear system equals
2.4.7 Time-Invariant Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Systems that don't change their input-output relation with time are said to be timeinvariant. The mathematical way of stating this property is to use the signal delay
concept described in Delay (Page 28).
If you delay (or advance) the input, the output is similarly delayed (advanced). Thus, a
time-invariant system responds to an input you may supply tomorrow the same way it
responds to the same input applied today; today's output is merely delayed to occur
tomorrow.
The collection of linear, time-invariant systems are the most thoroughly understood
systems. Much of the signal processing and system theory discussed here
concentrates on such systems. For example, electric circuits are, for the most part,
linear and time-invariant. Nonlinear ones abound, but characterizing them so that you
can predict their behavior for any input remains an unsolved problem.
31
Input-Output Relation
Linear
Time-Invariant
yes
yes
yes
yes
no
yes
yes
yes
yes
yes
yes
yes
yes
no
yes
no
no
yes
no
yes
no
yes
Table 2.1 Linear, Time-Invariant Table
32
2.5 Signals and Systems Problems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Problem 2.1: Complex Number
Arithmetic
Find the real part, imaginary part, the magnitude and angle of the
complex numbers given by the following expressions.
1.
2.
3.
4.
Problem 2.2: Discovering Roots
Complex numbers expose all the roots of real (and complex)
numbers. For example, there should be two square-roots, three
cube-roots, etc. of any number. Find the following roots.
1. What are the cube-roots of 27? In other words, what is
?
2. What are the fifth roots of
3. What are the fourth roots of one?
Problem 2.3: Cool Exponentials
Simplify the following (cool) expressions.
1.
2.
3.
33
Problem 2.4: Complex-valued Signals
Complex numbers and phasors play a very important role in
electrical engineering. Solving systems for complex exponentials is
much easier than for sinusoids, and linear systems analysis is
particularly easy.
1. Find the phasor representation for each, and re-express each as
the real and imaginary parts of a complex exponential. What is the
frequency (in Hz) of each? In general, are your answers unique? If
so, prove it; if not, find an alternative answer for the complex
exponential representation.
1.
2.
3.
2. Show that for linear systems having real-valued outputs for real
inputs, that when the input is the real part of a complex
exponential, the output is the real part of the system's output to
the complex exponential (see the below figure).
34
Problem 2.5:
For each of the indicated voltages, write it as the real part of a
complex exponential (v (t)= Re (Vest)). Explicitly indicate the value of
the complex amplitude V and the complex frequency s. Represent
each complex amplitude as a vector in the V -plane, and indicate the
location of the frequencies in the complex s-plane.
35
Problem 2.6:
Express each of the following signals (signal a-e) as a linear
combination of delayed and weighted step functions and ramps (the
integral of a step).
Figure 2.16 Signal (a)
Figure 2.17 signal (b)
Figure 2.18 signal (c)
Figure 2.19 signal (d)
36
Figure 2.20 signal (e)
Problem 2.7: Linear, Time-Invariant
Systems
When the input to a linear, time-invariant system is the signal x (t),
the output is the signal y (t) (The following figure).
1. Find and sketch this system's output when the input is the
depicted signal (Figure 2.19).
2. Find and sketch this system's output when the input is a unit step.
Figure 2.21
↵
37
Problem 2.8: Linear Systems
The depicted input (Figure below) x (t) to a linear, time-invariant
system yields the output y (t).
1. What is the system's output to a unit step input u (t)?
2. What will the output be when the input is the depicted square
wave in the following figure?
Problem 2.9: Communication Channel
A particularly interesting communication channel can be modeled as
a linear, time-invariant system. When the transmitted signal x (t) is a
pulse, the received signal r (t) is as shown (in the following Figure).
1. What will be the received signal when the transmitter sends the
pulse sequence (Figure 2.23) x1 (t)?
2. What will be the received signal when the transmitter sends the
pulse signal (the figure below) x2 (t) that has half the duration as
the original?
38
Problem 2.10: Analog Computers
So-called analog computers use circuits to solve mathematical
problems, particularly when they involve differential equations.
Suppose we are given the following differential equation to solve.
In this equation,a is a constant.
1. When the input is a unit step (x (t)= u (t)), the output is given by y
(t)= 1 − eu (t-(at)) u(t) . What is the total energy expended by the
input?
2. Instead of a unit step, suppose the input is a unit pulse (unitamplitude, unit-duration) delivered to the circuit at time t = 10.
What is the output voltage in this case? Sketch the waveform.
2.6 Solutions to Exercises in Chapter 2
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 2.1.1
z + z* = a + jb + a − jb =2a =2Re (z). Similarly, z – z* = a + jb − (a − jb)=2jb = 2jIm (z)
Solution to Exercise 2.1.2
To convert 3 – 2j to polar form, we first locate the number in the complex plane in the
fourth quadrant. The distance from the origin to the complex number is the
magnitude r, which equals
The angle equals
or −0.588 radians (−33.7 degrees). The final answer is
degrees.
Solution to Exercise 2.1.3
2
2
2
2
zz* =(a + jb)(a − jb)= a + b . Thus, zz* = r =(|z|) .
Solution to Exercise 2.3.1
Solution to Exercise 2.6.1
In the first case, order does not matter; in the second it does. "Delay" means t → t − τ.
"Time-reverse" means t →−t
39
Case 1 y (t)= Gx (t −τ), and the way we apply the gain and delay the signal gives the
same result.
Case 2 Time-reverse then delay: y (t)= x (− (t −τ )) = x (−t + τ). Delay then time-reverse: y
(t)= x ((−t) −τ ).
40
Chapter 3 Analog Signal Processing
3.1 Voltage, Current, and Generic Circuit Elements
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We know that information can be represented by signals; now we need to understand
how signals are physically realized. Over the years, electric signals have been found to
be the easiest to use. Voltage and currents comprise the electric instantiations of
signals. Thus, we need to delve into the world of electricity and electromagnetism. The
systems used to manipulate electric signals directly are called circuits, and they refine
the information representation or extract information from the voltage or current. In
many cases, they make nice examples of linear systems.
A generic circuit element places a constraint between the classic variables of a circuit:
voltage and current. Voltage is electric potential and represents the "push" that drives
electric charge from one place to another. What causes charge to move is a physical
separation between positive and negative charge. A battery generates, through
electrochemical means, excess positive charge at one terminal and negative charge at
the other, creating an electric field. Voltage is defined across a circuit element, with the
positive sign denoting a positive voltage drop across the element. When a conductor
connects the positive and negative potentials, current flows, with positive current
indicating that positive charge flows from the positive terminal to the negative.
Electrons comprise current flow in many cases. Because electrons have a negative
charge, electrons move in the opposite direction of positive current flow: Negative
charge flowing to the right is equivalent to positive charge moving to the left.
It is important to understand the physics of current flow in conductors to appreciate
the innovation of new electronic devices. Electric charge can arise from many sources,
the simplest being the electron. When we say that "electrons flow through a
conductor," what we mean is that the conductor's constituent atoms freely give up
electrons from their outer shells. "Flow" thus means that electrons hop from atom to
atom driven along by the applied electric potential. A missing electron, however, is a
virtual positive charge. Electrical engineers call these holes, and in some materials,
particularly certain semiconductors, current flow is actually due to holes. Current flow
also occurs in nerve cells found in your brain. Here, neurons "communicate" using
propagating voltage pulses that rely on the flow of positive ions (potassium and
sodium primarily, and to some degree calcium) across the neuron's outer wall. Thus,
current can come from many sources, and circuit theory can be used to understand
how current flows in reaction to electric fields.
Generic Circuit Element
41
Figure 3.1 The generic circuit element.
Current flows through circuit elements, such as that depicted in Figure 3.1, and
through conductors, which we indicate by lines in circuit diagrams. For every circuit
element we define a voltage and a current. The element has a v-i relation defined by
the element's physical properties. In defining the v-i relation, we have the convention
that positive current flows from positive to negative voltage drop. Voltage has units of
volts, and both the unit and the quantity are named for Volta. Current has units of
amperes, and is named for the French physicist Ampere.
Voltages and currents also carry power. Again using the convention shown in Figure
3.1 (Generic Circuit Element) for circuit elements, the instantaneous power at each
moment of time consumed by the element is given by the product of the voltage and
current.
A positive value for power indicates that at time t the circuit element is consuming
power; a negative value means it is producing power. With voltage expressed in volts
and current in amperes, power defined this way has units of watts. Just as in all areas
of physics and chemistry, power is the rate at which energy is consumed or produced.
Consequently, energy is the integral of power.
Again, positive energy corresponds to consumed energy and negative energy
corresponds to energy produc tion. Note that a circuit element having a power profile
that is both positive and negative over some time interval could consume or produce
energy according to the sign of the integral of power. The units of energy are joules
since a watt equals joules/second.
Exercise 3.1.1
Residential energy bills typically state a home's energy usage in
kilowatt-hours. Is this really a unit of energy? If so, how many joules
equals one kilowatt-hour?
3.2 Ideal Circuit Elements
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The elementary circuit elements the resistor, capacitor, and inductor impose linear
relationships between voltage and current.
42
3.2.1 Resistor
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.2 Resistor. v = Ri
The resistor is far and away the simplest circuit element. In a resistor, the voltage is
proportional to the current, with the constant of proportionality R, known as the
resistance.
Resistance has units of ohms, denoted by Ω, named for the German electrical scientist
Georg Ohm5 . Sometimes, the v-i relation for the resistor is written i = Gv, with G, the
conductance, equal to
. Conductance has units of Siemens (S), and is named for the German electronics
industrialist Werner von Siemens .
When resistance is positive, as it is in most cases, a resistor consumes power. A
resistor's instantaneous power consumption can be written one of two ways.
As the resistance approaches infinity, we have what is known as an open circuit: No
current flows but a non-zero voltage can appear across the open circuit. As the
resistance becomes zero, the voltage goes to zero for a non-zero current flow. This
situation corresponds to a short circuit. A superconductor physically realizes a short
circuit.
3.2.2 Capacitor
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.3 Capacitor.
The capacitor stores charge and the relationship between the charge stored and the
resultant voltage is q = Cv. The constant of proportionality, the capacitance, has units
of farads (F), and is named for the English experimental physicist Michae l Faraday. As
43
current is the rate of change of charge, the v-i relation can be expressed in differential
or integral form.
If the voltage across a capacitor is constant, then the current flowing into it equals
zero. In this situation, the capacitor is equivalent to an open circuit. The power
consumed/produced by a voltage applied to a capacitor depends on the product of
the voltage and its derivative.
This result means that a capacitor's total energy expenditure up to time t is concisely
given by
This expression presumes the fundamental assumption of circuit theory: all
voltages and currents in any circuit were zero in the far distant past (t = −∞).
3.2.3 Inductor
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.4 Inductor
The inductor stores magnetic flux, with larger valued inductors capable of storing
more flux. Inductance has units of henries (H), and is named for the American
physicist Joseph Henry8 . The differential and integral forms of the inductor's v-i
relation are
The power consumed/produced by an inductor depends on the product of the
inductor current and its derivative
and its total energy expenditure up to time t is given by
44
3.2.4 Sources
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.5 Sources The voltage source on the left and current source on the right are like all circuit
elements in that they have a particular relationship between the voltage and current defined for them. For
the voltage source, v = vs for any current i; for the current source, i = −is for any voltage v
Sources of voltage and current are also circuit elements, but they are not linear in the
strict sense of linear systems. For example, the voltage source's v-i relation is v = vs
regardless of what the current might be. As for the current source, i = −is regardless of
the voltage. Another name for a constant-valued voltage source is a battery, and can
be purchased in any supermarket. Current sources, on the other hand, are much
harder to acquire; we'll learn why later.
3.3 Ideal and Real-World Circuit Elements
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Source and linear circuit elements are ideal circuit elements. One central notion of
circuit theory is combining the ideal elements to describe how physical elements
operate in the real world. For example, the 1 kΩ resistor you can hold in your hand is
not exactly an ideal 1 kΩ resistor. First of all, physical devices are manufactured to
close tolerances (the tighter the tolerance, the more money you pay), but never have
exactly their advertised values. The fourth band on resistors specifes their tolerance;
10% is common. More pertinent to the current discussion is another deviation from
the ideal: If a sinusoidal voltage is placed across a physical resistor, the current will not
be exactly proportional to it as frequency becomes high, say above 1 MHz. At very
high frequencies, the way the resistor is constructed introduces inductance and
capacitance efects. Thus, the smart engineer must be aware of the frequency ranges
over which his ideal models match reality well.
On the other hand, physical circuit elements can be readily found that well
approximate the ideal, but they will always deviate from the ideal in some way. For
example, a fashlight battery, like a C-cell, roughly corresponds to a 1.5 V voltage
source. However, it ceases to be modeled by a voltage source capable of supplying
any current (that's what ideal ones can do!) when the resistance of the light bulb is too
small.
45
3.4 Electric Circuits and Interconnection Laws
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A circuit connects circuit elements together in a specific configuration designed to
transform the source signal (originating from a voltage or current source) into another
signal the output that corresponds to the current or voltage defined for a particular
circuit element. A simple resistive circuit is shown in Figure 3.6.
This circuit is the electrical embodiment of a system having its input provided by a
source system producing vin (t).
Figure 3.6 Circuits The circuit shown in the top two figures is perhaps the simplest circuit that performs a
signal processing function. On the bottom is the block diagram that corresponds to the circuit. The input is
provided by the voltage source vin and the output is the voltage v out across the resistor label R2. As shown
in the middle, we analyze the circuit understand what it accomplishes by defining currents and voltages for
all circuit elements, and then solving the circuit and element equations.
To understand what this circuit accomplishes, we want to determine the voltage
across the resistor labeled by its value R2. Recasting this problem mathematically, we
need to solve some set of equations so that we relate the output voltage vout to the
source voltage. It would be simple a little too simple at this point if we could instantly
write down the one equation that relates these two voltages. Until we have more
knowledge about how circuits work, we must write a set of equations that allow us to
find all the voltages and currents that can be defined for every circuit element.
46
Because we have a three-element circuit, we have a total of six voltages and currents
that must be either specified or determined. You can define the directions for positive
current flow and positive voltage drop any way you like. Once the values for the
voltages and currents are calculated, they may be positive or negative according to
your definition. When two people define variables according to their individual
preferences, the signs of their variables may not agree, but current flow and voltage
drop values for each element will agree. Do recall in defining your Ideal Circuit
Elements (Page 41) that the v-i relations for the elements presume that positive
current flow is in the same direction as positive voltage drop. Once you define
voltages and currents, we need six nonredundant equations to solve for the six
unknown voltages and currents. By specifying the source, we have one; this amounts
to providing the source's v-i relation. The v-i relations for the resistors give us two
more. We are only halfway there; where do we get the other three equations we
need?
What we need to solve every circuit problem are mathematical statements that
express how the circuit elements are interconnected. Said another way, we need the
laws that govern the electrical connection of circuit elements. First of all, the places
where circuit elements attach to each other are called nodes. Two nodes are explicitly
indicated in Figure 3.6; a third is at the bottom where the voltage source and resistor
R2are connected. Electrical engineers tend to draw circuit diagrams schematics in a
rectilinear fashion. Thus the long line connecting the bottom of the voltage source
with the bottom of the resistor is intended to make the diagram look pretty. This line
simply means that the two elements are connected together. Kirchhof's Current Law
(Page 46), one for voltage and one for current, determine what a connection among
circuit elements means. These laws are essential to analyzing this and any circuit. They
are named for Gustav Kirchhofll , a nineteenth century German physicist.
3.4.1 Kirchhof's Current Law
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
At every node, the sum of all currents entering or leaving a node must equal zero.
What this law means physically is that charge cannot accumulate in a node; what goes
in must come out. In the example, Figure 3.6, below we have a three-node circuit and
thus have three KCL equations.
Note that the current entering a node is the negative of the current leaving the node.
Given any two of these KCL equations, we can find the other by adding or subtracting
them. Thus, one of them is redundant and, in mathematical terms, we can discard any
one of them. The convention is to discard the equation for the (unlabeled) node at the
bottom of the circuit.
47
Figure 3.7 Simple circult The circuit shown is perhaps the simplest circuit that performs a signal
processing function. The input is provided by the voltage source v in and the output is the voltage vout across
the resistor labelled R2.
Exercise 3.4.1
In writing KCL equations, you will find that in an n-node circuit,
exactly one of them is always redundant. Can you sketch a proof of
why this might be true? Hint: It has to do with the fact that charge
won't accumulate in one place on its own.
3.4.2 Kirchhof's Voltage Law (KVL)
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The voltage law says that the sum of voltages around every closed loop in the circuit
must equal zero. A closed loop has the obvious defnition: Starting at a node, trace a
path through the circuit that returns you to the origin node. KVL expresses the fact
that electric fields are conservative: The total work performed in moving a test charge
around a closed path is zero. The KVL equation for our circuit is
In writing KVL equations, we follow the convention that an element's voltage enters
with a plus sign when traversing the closed path, we go from the positive to the
negative of the voltage's defnition.
For the example circuit (Simple circult), we have three v-i relations, two KCL equations,
and one KVL equation for solving for the circuit's six voltages and currents.
v-i:
KCL:
KVL:
48
We have exactly the right number of equations! Eventually, we will discover shortcuts
for solving circuit problems; for now, we want to eliminate all the variables but vout
and determine how it depends on vin and on resistor values. The KVL equation can be
rewritten as vin = v1 + vout . Substituting into it the resistor's v-i relation, we have vin =
R1i1 + R2iout . Yes, we temporarily eliminate the quantity we seek. Though not obvious,
it is the simplest way to solve the equations. One of the KCL equations says i1 = iout,
which means that vin = R1iout + R2iout =(R1 + R2) iout. Solving for the current in the
output resistor, we have
. We have now solved the circuit: We have expressed one voltage or current in
terms of R1+R2sources and circuit-element values. To find any other circuit quantities,
we can back substitute this answer into our original equations or ones we developed
along the way. Using the v-i relation for the output resistor, we obtain the quantity we
seek.
Exercise 3.4.2
Referring back to Figure 3.6, a circuit should serve some useful
purpose. What kind of system does our circuit realize and, in terms of
element values, what are the system's parameter(s)?
3.5 Power Dissipation in Resistor Circuits
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We can find voltages and currents in simple circuits containing resistors and voltage or
current sources. We should examine whether these circuits variables obey the
Conservation of Power principle: since a circuit is a closed system, it should not
dissipate or create energy. For the moment, our approach is to investigate first a
resistor circuit's power consumption/creation. Later, we will prove that because of
KVL and KCL all circuits conserve power.
As defined previously, the instantaneous power consumed/created by every circuit
element equals the product of its voltage and current. The total power consumed/
created by a circuit equals the sum of each element's power.
Recall that each element's current and voltage must obey the convention that positive
current is defined to enter the positive-voltage terminal. With this convention, a
positive value of vkik corresponds to consumed power, a negative value to created
power. Because the total power in a circuit must be zero (P =0), some circuit elements
must create power while others consume it.
49
Consider the simple series circuit should in Electric Circuits and Interconnection Laws
(Page 45). In performing our calculations, we defined the current iout to flow through
the positive-voltage terminals of both resistors and found it to equal
. The voltage across the resistor R2 is the output voltage and we found it to equal
Consequently, calculating the power for this resistor yields
Consequently, this resistor dissipates power because P2 is positive. This result should
not be surprising since we showed (p. 41) that the power consumed by any resistor
equals either of the following.
Since resistors are positive-valued, resistors always dissipate power. But where
does a resistor's power go? By Conservation of Power, the dissipated power must be
absorbed somewhere. The answer is not directly predicted by circuit theory, but is by
physics. Current fowing through a resistor makes it hot; its power is dissipated by
heat.
Note: A physical wire has a resistance and hence dissipates power (it gets
warm just like a resistor in a circuit). In fact, the resistance of a wire of length L
and cross-sectional area A is given by
The quantity ρ is known as the resistivity and presents the resistance of a unit-length,
unit cross-sectional area material constituting the wire. Resistivity has units of ohmmeters. Most materials have a positive value for ρ, which means the longer the wire,
the greater the resistance and thus the power dissipated. The thicker the wire, the
smaller the resistance. Superconductors have zero resistivity and hence do not
dissipate power. If a room-temperature superconductor could be found, electric
power could be sent through power lines without loss!
50
Exercise 3.5.1
Calculate the power consumed/created by the resistor R1 in our
simple circuit example. We conclude that both resistors in our
example circuit consume power, which points to the voltage source
as the producer of power. The current flowing into the source's
positive terminal is −iout . Consequently, the power calculation for
the source yields
We conclude that the source provides the power consumed by the
resistors, no more, no less.
Exercise 3.5.2
Confirm that the source produces exactly the total power consumed
by both resistors.
This result is quite general: sources produce power and the circuit
elements, especially resistors, consume it. But where do sources get
their power? Again, circuit theory does not model how sources are
constructed, but the theory decrees that all sources must be provided
energy to work.
3.6 Series and Parallel Circuits
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.8 Circult examples The circuit shown is perhaps the simplest circuit that performs a signal
processing function. The input is provided by the voltage source v in and the output is the voltage vout across
the resistor labelled R2
The results shown in other modules (Electric Circuits and Interconnection Laws (Page
45), KVL and KCL, interconnection laws) with regard to this circuit (Figure 3.8), and the
values of other currents and voltages in this circuit as well, have profound
implications.
Resistors connected in such a way that current from one must flow only into another
currents in all resistors connected this way have the same magnitude are said to be
51
connected in series. For the two series-connected resistors in the example, the
voltage across one resistor equals the ratio of that resistor's value and the sum
of resistances times the voltage across the series combination. This concept is so
pervasive it has a name: voltage divider.
The input-output relationship for this system, found in this particular case by
voltage divider, takes the form of a ratio of the output voltage to the input voltage.
In this way, we express how the components used to build the system affect the inputoutput relationship. Because this analysis was made with ideal circuit elements, we
might expect this relation to break down if the input amplitude is too high (Will the
circuit survive if the input changes from 1 volt to one million volts?) or if the source's
frequency becomes too high. In any case, this important way of expressing inputoutput relationships as a ratio of output to input pervades circuit and system theory.
The current i1 is the current flowing out of the voltage source. Because it equals i2, we
have that
RESISTORS IN SERIES: The series combination of two resistors acts, as far as the
voltage source is concerned, as a single resistor having a value equal to the sum of the
two resistances.
This result is the first of several equivalent circuit ideas: In many cases, a complicated
circuit when viewed from its terminals (the two places to which you might attach a
source) appears to be a single circuit element (at best) or a simple combination of
elements at worst. Thus, the equivalent circuit for a series combination of resistors is a
single resistor having a resistance equal to the sum of its component resistances.
Figure 3.9 Resistors The resistor (on the right) is equivalent to the two resistors (on the left) and has a
resistance equal to the sum of the resistances of the other two resistors.
Thus, the circuit the voltage source "feels" (through the current drawn from it) is a
single resistor having resistance R1 + R2. Note that in making this equivalent circuit,
the output voltage can no longer be defined: The output resistor labeled R2 no longer
appears. Thus, this equivalence is made strictly from the voltage source's viewpoint.
52
:
Figure 3.10 A simple parallel circuit.
One interesting simple circuit (Figure 3.10) has two resistors connected side-by-side,
what we will term a parallel connection, rather than in series. Here, applying KVL
reveals that all the voltages are identical: v1 = v and v2 = v. This result typifies parallel
connections. To write the KCL equation, note that the top node consists of the entire
upper interconnection section. The KCL equation is iin − i1 − i2 =0. Using the v-i
relations, we find that
Exercise 3.6.1
Suppose that you replaced the current source in Figure 3.10 by a
voltage source. How would iout be related to the source voltage? Based
on this result, what purpose does this revised circuit have? This
circuit highlights some important properties of parallel circuits. You
can easily show that the parallel combination of R1 and R2 has the v-i
relation of a resistor having resistance
. A notation for this quantity is ( R1 || R2). As the reciprocal of
resistance is conductance (Resistor (Page 42) ), we can say that for a
parallel combination of resistors, the equivalent conductance is
the sum of the conductances.
Figure 3.11 A parallel combination of resistors
Similar to voltage divider for series resistances, we have current divider for parallel
resistances. The current through a resistor in parallel with another is the ratio of the
conductance of the first to the sum of the conductances. Thus, for the depicted circuit,
53
. Expressed in terms of resistances, current divider takes the form of the resistance of
the other resistor divided by the sum of resistances:
Figure 3.12 Circult example 2
:
Figure 3.13 Circuit example 3 The simple attenuator circuit is attached to an oscilloscope's input. The
input-output relation for the above circuit without a load is: .
Suppose we want to pass the output signal into a voltage measurement device, such
as an oscilloscope or a voltmeter. In system-theory terms, we want to pass our circuit's
output to a sink. For most applications, we can represent these measurement devices
as a resistor, with the current passing through it driving the measurement device
through some type of display. In circuits, a sink is called a load; thus, we describe a
system-theoretic sink as a load resistance RL. Thus, we have a complete system built
from a cascade of three systems: a source, a signal processing system (simple as it is),
and a sink.
We must analyze afresh how this revised circuit, shown in Figure 3.13, works. Rather
than defining eight variables and solving for the current in the load resistor, let's take
a hint from other analysis (series rules, parallel rules). Resistors R2 and RL are in a
parallel configuration: The voltages across each resistor are the same while the
currents are not. Because the voltages are the same, we can find the current through
each from their v-i relations:
and
. Considering the node where all three resistors join, KCL says that the sum of the
three currents must equal zero. Said another way, the current entering the node
54
through R1 must equal the sum of the other two currents leaving the node. Therefore,
i1 = i2 + iL, which means that
Let Req denote the equivalent resistance of the parallel combination of R2 and RL.
Using R1'sv-i relation, the voltage across it is
. The KVL equation written around the leftmost loop has vin= v1 + vout ; substituting for
v1, we find
or
Thus, we have the input-output relationship for our entire system having the form of
voltage divider, but it does not equal the input-output relation of the circuit without
the voltage measurement device. We can not measure voltages reliably unless the
measurement device has little effect on what we are trying to measure. We should
look more carefully to determine if any values for the load resistance would lessen its
impact on the circuit. Comparing the input-output relations before and after, what we
need is
As
, the approximation would apply if
or
This is the condition we seek:
VOLTAGE MEASUREMENT: Voltage measurement devices must have large resistances
compared with that of the resistor across which the voltage is to be measured.
Exercise 3.6.2
Let's be more precise: How much larger would a load resistance need
to be to affect the input-output relation by less than 10%? by less
than 1%?
55
Example 3.1
Figure 3.14 Example 3.1
We want to find the total resistance of the example circuit. To apply
the series and parallel combination rules, it is best to first determine
the circuit's structure: What is in series with what and what is in
parallel with what at both small-and large-scale views. We have R2
in parallel with R3; this combination is in series with R4. This series
combination is in parallel with R1. Note that in determining this
structure, we started away from the terminals, and worked toward
them. In most cases, this approach works well; try it first. The total
resistance expression mimics the structure:
Such complicated expressions typify circuit "simplifications." A
simple check for accuracy is the units: Each component of the
numerator should have the same units (here Ω3) as well as in the
denominator (Ω2). The entire expression is to have units of
resistance; thus, the ratio of the numerator's and denominator's
units should be ohms. Checking units does not guarantee accuracy,
but can catch many errors.
Another valuable lesson emerges from this example concerning the difference
between cascading systems and cascading circuits. In system theory, systems can be
cascaded without changing the input-output relation of intermediate systems. In
cascading circuits, this ideal is rarely true unless the circuits are so designed. Design
is in the hands of the engineer; he or she must recognize what have come to be
known as loading effects. In our simple circuit, you might think that making the
resistance RL large enough would do the trick. Because the resistors R1 and R2 can
have virtually any value, you can never make the resistance of your voltage
measurement device big enough. Said another way, a circuit cannot be designed in
isolation that will work in cascade with all other circuits. Electrical engineers deal
with this situation through the notion of specifications: Under what conditions will
the circuit perform as designed? Thus, you will find that oscilloscopes and voltmeters
have their internal resistances clearly stated, enabling you to determine whether the
56
voltage you measure closely equals what was present before they were attached to
your circuit. Furthermore, since our resistor circuit functions as an attenuator, with the
attenuation (a fancy word for gains less than one) depending only on the ratio of the
two resist or values
, we can select any values for the two resistances we want to achieve the desired
attenuation. The designer of this circuit must thus specify not only what the
attenuation is, but also the resistance values employed so that integrators people who
put systems together from component systems can combine systems together and
have a chance of the combination working.
Figure 3.15 summarizes the series and parallel combination results. These results are
easy to remember and very useful. Keep in mind that for series combinations, voltage
and resistance are the key quantities, while for parallel combinations current and
conductance are more important. In series combinations, the currents through each
element are the same; in parallel ones, the voltages are the same.
Figure 3.15 Series and parallel combonation rules
(a)
(b)
Exercise 3.6.3
Contrast a series combination of resistors with a parallel one. Which
variable (voltage or current) is the same for each and which differs?
What are the equivalent resistances? When resistors are placed in
series, is the equivalent resistance bigger, in between, or smaller
than the component resistances? What is this relationship for a
parallel combination?
57
3.7 Equivalent Circuits: Resistors and Sources
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We have found that the way to think about circuits is to locate and group parallel and
series resistor combinations. Those resistors not involved with variables of interest
can be collapsed into a single resistance. This result is known as an equivalent
circuit: from the viewpoint of a pair of terminals, a group of resistors functions as a
single resistor, the resistance of which can usually be found by applying the parallel
and series rules.
This result generalizes to include sources in a very interesting and useful way. Let's
consider our simple attenuator circuit (shown in the figure (Figure 3.16) from the
viewpoint of the output terminals. We want to find the v-i relation for the output
terminal pair, and then find the equivalent circuit for the boxed circuit. Two perform
this calculation, use the circuit laws and element relations, but do not attach anything
to the output terminals. We seek the relation between v and i that describes the kind
of element that lurks within the dashed box. The result is
Figure 3.16 Simple attenuator circuit
If the source were zero, it could be replaced by a short circuit, which would confirm
that the circuit does indeed function as a parallel combination of resistors. However,
the source's presence means that the circuit is not well modeled as a resistor.
Figure 3.17 The Thevenin equivalent circult
58
If we consider the simple circuit of Figure 3.17, we find it has the v-i relation at its
terminals of
Comparing the two v-i relations, we find that they have the same form. In this case the
Thevenin equivalent resistance is Req = ( R1 || R2 ) and the Thevenin equivalent
source has voltage
Thus, from viewpoint of the terminals, you cannot distinguish the two circuits.
Because the equivalent circuit has fewer elements, it is easier to analyze and
understand than any other alternative.
For any circuit containing resistors and sources, the v-i relation will be of the form
and the Thevenin equivalent circuit for any such circuit is that of Figure 3.17. This
equivalence applies no matter how many sources or resistors may be present in the
circuit. In the example (Example 3.2) below, we know the circuit's construction and
element values, and derive the equivalent source and resistance. Because Thevenin's
theorem applies in general, we should be able to make measurements or calculations
only from the terminals to determine the equivalent circuit.
To be more specific, consider the equivalent circuit of this figure (Figure 3.17). Let the
terminals be opencircuited, which has the effect of setting the current i to zero.
Because no current flows through the resistor, the voltage across it is zero (remember,
Ohm's Law says that v = Ri). Consequently, by applying KVL we have that the so-called
open-circuit voltage voc equals the Thevenin equivalent voltage. Now consider the
situation when we set the terminal voltage to zero (short-circuit it) and measure the
resulting current. Referring to the equivalent circuit, the source voltage now appears
entirely across the resistor, leaving the short-circuit current to be
. From this property, we can determine the equivalent resistance.
59
Exercise 3.7.1
Use the open/short-circuit approach to derive the Thevenin
equivalent of the circuit shown in Figure 3.18.
Figure 3.18 Circuit for exercise
Example 3.2
Figure 3.19 Circuit sample
For the circuit depicted in Figure 3.19, let's derive its Thevenin
equivalent two different ways. Starting with the open/short-circuit
approach, let's first find the open-circuit voltage voc. We have a
current divider relationship as R1 is in parallel with the series
combination of R2 and R3.Thus,
When we short-circuit the terminals, no voltage appears across R3,
and thus no current flows through it. In short, R3 does not affect the
short-circuit current, and can be eliminated. We again have a current
divider relationship:
. Thus, the Thevenin equivalent resistance is
To verify, let's find the equivalent resistance by reaching inside the circuit and setting
the current source to zero. Because the current is now zero, we can replace the
current source by an open circuit. From the viewpoint of the terminals, resistor R3 is
now in parallel with the series combination of R1 and R2. Thus, Req = ( R3 || R1 + R2),
and we obtain the same result.
60
Figure 3.20 Thevenin and Mayer-Norton equivalent All circuits containing sources and resistors can be
described by simpler equivalent circuits. Choosing the one to use depends on the application, not on what is
actually inside the circuit.
As you might expect, equivalent circuits come in two forms: the voltage-source
oriented Thevenin equivalentl5 and the current-source oriented Mayer-Norton
equivalent (Figure 3.20). To derive the latter, the v-i relation for the Thevenin
equivalent can be written as
or
where
is the Mayer-Norton equivalent source. The Mayer-Norton equivalent shown in Figure
3.20 can be easily shown to have this v-i relation. Note that both variations have the
same equivalent resistance. The short-circuit current equals the negative of the
Mayer-Norton equivalent source.
61
Exercise 3.7.2
Find the Mayer-Norton equivalent circuit for the circuit below.
Figure 3.21 Circuit for exercise
Equivalent circuits can be used in two basic ways. The first is to
simplify the analysis of a complicated circuit by realizing the any
portion of a circuit can be described by either a Thevenin or MayerNorton equivalent. Which one is used depends on whether what is
attached to the terminals is a series confguration (making the
Thevenin equivalent the best) or a parallel one (making MayerNorton the best).
Another application is modeling. When we buy a fashlight battery,
either equivalent circuit can accurately describe it. These models help
us understand the limitations of a battery. Since batteries are labeled
with a voltage specifcation, they should serve as voltage sources and
the Thevenin equivalent serves as the natural choice. If a load
resistance RL is placed across its terminals, the voltage output can be
found using voltage divider :
If we have a load resistance much larger than the battery's equivalent
resistance, then, to a good approximation, the battery does serve as a
voltage source. If the load resistance is much smaller, we certainly
don't have a voltage source (the output voltage depends directly on
the load resistance). Consider now the Mayer-Norton equivalent; the
current through the load resistance is given by current divider, and
equals
For a current that does not vary with the load resistance, this
resistance should be much smaller than the equivalent resistance. If
the load resistance is comparable to the equivalent resistance, the
battery serves neither as a voltage source or a current course. Thus,
when you buy a battery, you get a voltage source if its equivalent
resistance is much smaller than the equivalent resistance of the
circuit to which you attach it. On the other hand, if you attach it to a
circuit having a small equivalent resistance, you bought a current
source.
62
LEON CHARLES THEVENIN: He was an engineer with France's Postes, Telegraphe et
Telephone. In 1883, he published (twice!) a proof of what is now called the Thevenin
equivalent while developing ways of teaching electrical engineering concepts at the
Ecole Polytechnique. He did not realize that the same result had been published by
Hermann Helmholtz, the renowned nineteenth century physicist, thiry years earlier.
HANS FERDINAND MAYER: After earning his doctorate in physics in 1920, he turned to
communications engineering when he joined Siemens & Halske in 1922. In 1926, he
published in a German technical journal the Mayer-Norton equivalent. During his
interesting career, he rose to lead Siemen's Central Laboratory in 1936, surruptiously
leaked to the British all he knew of German warfare capabilities a month after the
Nazis invaded Poland, was arrested by the Gestapo in 1943 for listening to BBC radio
broadcasts, spent two years in Nazi concentration camps, and went to the United
States for four years working for the Air Force and Cornell University before returning
to Siemens in 1950. He rose to a position on Siemen's Board of Directors before
retiring.
EDWARD L. NORTON: Edward Norton was an electrical engineer who worked at Bell
Laboratory from its inception in 1922. In the same month when Mayer's paper
appeared, Norton wrote in an internal technical memorandum a paragraph describing
the current-source equivalent. No evidence suggests Norton knew of Mayer's
publication.
3.8 Circuits with Capacitors and Inductors
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.22 A simple RC circuit.
Let's consider a circuit having something other than resistors and sources. Because of
KVL, we know that vin= vR+ vout. The current through the capacitor is given by
, and this current equals that passing through the resistor. Substituting vR = Ri into the
KVL equation and using the v-i relation for the capacitor, we arrive at
The input-output relation for circuits involving energy storage elements takes the form
of an ordinary differential equation, which we must solve to determine what the
output voltage is for a given input. In contrast to resistive circuits, where we obtain an
explicit input-output relation, we now have an implicit relation that requires more
work to obtain answers.
63
At this point, we could learn how to solve differential equations. Note first that even
finding the differential equation relating an output variable to a source is often very
tedious. The parallel and series combination rules that apply to resistors don't directly
apply when capacitors and inductors occur. We would have to slog our way through
the circuit equations, simplifying them until we fnally found the equation that related
the source(s) to the output. At the turn of the twentieth century, a method was
discovered that not only made fnding the differential equation easy, but also simplifed
the solution process in the most common situation. Although not original with him,
Charles Steinmetzl9 presented the key paper describing the impedance approach in
1893. It allows circuits containing capacitors and inductors to be solved with the same
methods we have learned to solved resistor circuits. To use impedances, we must
master complex numbers. Though the arithmetic of complex numbers is
mathematically more complicated than with real numbers, the increased insight into
circuit behavior and the ease with which circuits are solved with impedances is well
worth the diversion. But more importantly, the impedance concept is central to
engineering and physics, having a reach far beyond just circuits.
3.9 The Impedance Concept
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Rather than solving the differential equation that arises in circuits containing
capacitors and inductors, let's pretend that all sources in the circuit are complex
exponentials having the same frequency. Although this pretense can only be
mathematically true, this fction will greatly ease solving the circuit no matter what the
source really is.
Figure 3.23 A simple RC circuit
For the above example RC circuit (Figure 3.23), let
The complex amplitudeVin determines the size of the source and its phase. The critical
consequence of assuming that sources have this form is that all voltages and currents
in the circuit are also complex exponentials, having amplitudes governed by KVL, KCL,
and the v-i relations and the same frequency as the source. To appreciate why this
should be true, let's investigate how each circuit element behaves when either the
voltage or current is a complex exponential. For the resistor, v = Ri. When
; then
64
Thus, if the resistor's voltage is a complex exponential, so is the current, with an
amplitude
(determined by the resistor's v-i relation) and a frequency the same as the voltage.
Clearly, if the current were assumed to be a complex exponential, so would the
voltage. For a capacitor,
Letting the voltage be a complex exponential, we have
The amplitude of this complex exponential is
Finally, for the inductor, where
assuming the current to be a complex exponential results in the voltage having the
form
making its complex amplitude
.
The major consequence of assuming complex exponential voltage and currents
is that the ratio
for each element does not depend on time, but does depend on source
frequency. This quantity is known as the element's impedance.
Impedance
Figure 3.24 Independence
(a) Resistor:
(b) Capacitor:
(c) Inductor:
65
The impedance is, in general, a complex-valued, frequency-dependent quantity. For
example, the magnitude of the capacitor's impedance is inversely related to
frequency, and has a phase of
This observation means that if the current is a complex exponential and has constant
amplitude, the amplitude of the voltage decreases with frequency.
Let's consider Kirchof's circuit laws. When voltages around a loop are all complex
exponentials of the same frequency, we have
which means
the complex amplitudes of the voltages obey KVL. We can easily imagine that the
complex amplitudes of the currents obey KCL.
What we have discovered is that source(s) equaling a complex exponential of the
same frequency forces all circuit variables to be complex exponentials of the same
frequency. Consequently, the ratio of voltage to current for each element equals the
ratio of their complex amplitudes, which depends only on the source's frequency and
element values.
This situation occurs because the circuit elements are linear and time-invariant. For
example, suppose we had a circuit element where the voltage equaled the square of
the current:
meaning that voltage and current no longer had the same frequency and that their
ratio was time-dependent.
Because for linear circuit elements the complex amplitude of voltage is proportional
to the complex amplitude of current V = ZI assuming complex exponential sources
means circuit elements behave as if they were resistors, where instead of resistance,
we use impedance. Because complex amplitudes for voltage and current also
obey Kirchoff's laws, we can solve circuits using voltage and current divider and
the series and parallel combination rules by considering the elements to be
impedances.
3.10 Time and Frequency Domains
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we find the differential equation relating the source and the output, we are
faced with solving the circuit in what is known as the time domain. What we
66
emphasize here is that it is often easier to find the output if we use impedances.
Because impedances depend only on frequency, we find ourselves in the frequency
domain. A common error in using impedances is keeping the time-dependent part,
the complex exponential, in the fray. The entire point of using impedances is to get rid
of time and concentrate on frequency. Only after we find the result in the frequency
domain do we go back to the time domain and put things back together again.
To illustrate how the time domain, the frequency domain and impedances ft
together, consider the time domain and frequency domain to be two work rooms.
Since you can't be two places at the same time, you are faced with solving your circuit
problem in one of the two rooms at any point in time. Impedances and complex
exponentials are the way you get between the two rooms. Security guards make sure
you don't try to sneak time domain variables into the frequency domain room and
vice versa. Figure 3.25 (Two Rooms) shows how this works.
Two Rooms
Figure 3.25 Two Rooms The time and frequency domains are linked by assuming signals are complex
exponentials. In the time domain, signals can have any form. Passing into the frequency domain "work
room," signals are represented entirely by complex amplitudes.
As we unfold the impedance story, we'll see that the powerful use of impedances
suggested by Steinmetz greatly simplifies solving circuits, alleviates us from solving
differential equations, and suggests a general way of thinking about circuits. Because
of the importance of this approach, let's go over how it works.
1. Even though it's not, pretend the source is a complex exponential. We do this
because the impedance approach simplifies finding how input and output are
related. If it were a voltage source having voltage vin = p (t) (a pulse), still let
We'll learn how to "get the pulse back" later.
2. With a source equaling a complex exponential, all variables in a linear circuit will
also be complex exponentials having the same frequency. The circuit's only
remaining "mystery" is what each variable's complex amplitude might be. To find
67
these, we consider the source to be a complex number (Vin here) and the
elements to be impedances.
3. We can now solve using series and parallel combination rules how the complex
amplitude of any variable relates to the sources complex amplitude.
Example 3.3
To illustrate the impedance approach, we refer to the RC circuit
(Figure 3.26 (Simple Circuits)) below, and we assume that
.
Simple Circuits
Figure 3.26 Simple Circuits (a) A simple RC circuit. (b) The impedance counterpart for the RC circuit. Note
that the source and output voltage are now complex amplitudes.
Using impedances, the complex amplitude of the output voltage Vout can be found
using voltage divider:
If we refer to the differential equation for this circuit (shown in Circuits with Capacitors
and Inductors (Section 3.8) to be
, letting the output and input voltages be complex exponentials, we obtain the same
relationship between their complex amplitudes. Thus, using impedances is equivalent
to using the differential equation and solving it when the source is a complex
exponential.
In fact, we can find the differential equation directly using impedances. If we crossmultiply the relation between input and output amplitudes,
68
and then put the complex exponentials back in, we have
In the process of defining impedances, note that the factor j2πf arises from the
derivative of a complex exponential. We can reverse the impedance process, and
revert back to the differential equation.
This is the same equation that was derived much more tediously in Circuits with
Capacitors and Inductors (Section 3.8). Finding the differential equation relating
output to input is far simpler when we use impedances than with any other technique.
Exercise 3.10.1
Suppose you had an expression where a complex amplitude was
divided by j2πf. What time-domain operation corresponds to this
division?
3.11 Power in the Frequency Domain
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Recalling that the instantaneous power consumed by a circuit element or an
equivalent circuit that represents a collection of elements equals the voltage times the
current entering the positive-voltage terminal, p (t)= v (t) i (t), what is the equivalent
expression using impedances? The resulting calculation reveals more about power
consumption in circuits and the introduction of the concept of average power.
When all sources produce sinusoids of frequency f, the voltage and current for any
circuit element or collection of elements are sinusoids of the same frequency.
Here, the complex amplitude of the voltage V equals |V |e
jθ
jφ
and that of the current is
|I|e . We can also write the voltage and current in terms of their complex amplitudes
using Euler's formula (Euler's Formula (Page 13)).
Multiplying these two expressions and simplifying gives
69
We define
to be complex power. The real-part of complex power is the first term and since it
does not change with time, it represents the power consistently consumed/produced
by the circuit. The second term varies with time at a frequency twice that of the
source. Conceptually, this term details how power "sloshes" back and forth in the
circuit because of the sinusoidal source.
From another viewpoint, the real-part of complex power represents long-term energy
consumption/production. Energy is the integral of power and, as the integration
interval increases, the first term appreciates while the time-varying term "sloshes."
Consequently, the most convenient Definition of the average power consumed/
produced by any circuit is in terms of complex amplitudes.
Exercise 3.11.1
Suppose the complex amplitudes of the voltage and current have
fixed magnitudes. What phase relationship between voltage and
current maximizes the average power? In other words, how are φ and
θ related for maximum power dissipation?
Because the complex amplitudes of the voltage and current are
related by the equivalent impedance, average power can also be
written as
These expressions generalize the results (3.3) we obtained for resistor circuits. We
have derived a fundamental result: Only the real part of impedance contributes to
long-term power dissipation. Of the circuit elements, only the resistor dissipates
power. Capacitors and inductors dissipate no power in the long term. It is important to
realize that these statements apply only for sinusoidal sources. If you turn on a
constant voltage source in an RC-circuit, charging the capacitor does consume power.
70
Exercise 3.11.2
In an earlier problem (Introduction Problems (Page 9) ), we found
that the rms value of a sinusoid was
its amplitude divided by 2. What is average power expressed in terms
of the rms values of the voltage and current (Vrms and Irms
respectively)?
3.12 Equivalent Circuits: Impedances and Sources
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we have circuits with capacitors and/or inductors as well as resistors and
sources, Thevenin and Mayer-Norton equivalent circuits can still be defined by using
impedances and complex amplitudes for voltage and currents. For any circuit
containing sources, resistors, capacitors, and inductors, the input-output relation for
the complex amplitudes of the terminal voltage and current is
with Veq = ZeqIeq. Thus, we have Thevenin and Mayer-Norton equivalent circuits as
shown in Figure 3.27.
Equivalent Circuits
71
Figure 3.27 Equivalent Circuits Comparing the first, simpler, figure with the slightly more complicated
second figure, we see two differences. First of all, more circuits (all those containing linear elements in fact)
have equivalent circuits that contain equivalents. Secondly, the terminal and source variables are now
complex amplitudes, which carries the implicit assumption that the voltages and currents are single
complex exponentials, all having the same frequency.
72
Example 3.4
Simple RC Circuit
Figure 3.28 Simple RC Circuit
Let's find the Thevenin and Mayer-Norton equivalent circuits for
Figure 3.28 (Simple RC Circuit). The open-circuit voltage and shortcircuit current techniques still work, except we use impedances and
complex amplitudes. The open-circuit voltage corresponds to the
transfer function we have already found. When we short the
terminals, the capacitor no longer has any effect on the circuit, and
the short-circuit current Isc equals
. The equivalent impedance can be found by setting the source to
zero, and finding the impedance using series and parallel
combination rules. In our case, the resistor and capacitor are in
parallel once the voltage source is removed (setting it to zero
amounts to replacing it with a short-circuit). Thus,
Consequently, we have
Again, we should check the units of our answer. Note in particular
that j2πfRC must be dimensionless. Is it?
73
3.13 Transfer Functions
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The ratio of the output and input amplitudes for Figure 3.29 ,known as the transfer
function or the frequency response, is given by
Implicit in using the transfer function is that the input is a complex exponential, and
the output is also a complex exponential having the same frequency. The transfer
function reveals how the circuit modifies the input amplitude in creating the output
amplitude. Thus, the transfer function completely describes how the circuit processes
the input complex exponential to produce the output complex exponential. The
circuit's function is thus summarized by the transfer function. In fact, circuits are often
designed to meet transfer function specifications. Because transfer functions are
complex-valued, frequency-dependent quantities, we can better appreciate a circuit's
function by examining the magnitude and phase of its transfer function (Figure 3.30
(Magnitude and phase of the transfer function)).
Simple Circuit
Figure 3.29 A simple RC circuit
74
Figure 3.30 Magnitude and phase of the transfer function (Simple Circuit) when RC=1.
(a)
(b)
This transfer function has many important properties and provides all the insights
needed to determine how the circuit functions. First of all, note that we can compute
the frequency response for both positive and negative frequencies. Recall that
sinusoids consist of the sum of two complex exponentials, one having the negative
frequency of the other. We will consider how the circuit acts on a sinusoid soon. Do
note that the magnitude has even symmetry: The negative frequency portion is a
mirror image of the positive frequency portion: |H (−f) | = |H (f) |. The phase has odd
symmetry: ∠ (H (−f)) = −(∠ (H (f)). These properties of this specific example apply for all
transfer functions associated with circuits. Consequently, we don't need to plot the
negative frequency component; we know what it is from the positive frequency part.
The magnitude equals
of its maximum gain (1 at f =0) when 2πfRC =1 (the two terms in the 2 denominator of
the magnitude are equal). The frequency
defines the boundary between two operating ranges.
• For frequencies below this frequency, the circuit does not much alter the
amplitude of the complex exponential source.
75
• For frequencies greater than fc, the circuit strongly attenuates the amplitude.
Thus, when the source frequency is in this range, the circuit's output has a much
smaller amplitude than that of the source.
For these reasons, this frequency is known as the cutoff frequency. In this circuit the
cutoff frequency depends only on the product of the resistance and the capacitance.
Thus, a cutoff frequency of 1 kHz occurs when
or
. Thus resistance-capacitance combinations of 1.59 kΩ and 100nF or 10 Ω and 1.59 µF
result in the same cutoff frequency.
The phase shift caused by the circuit at the cutoff frequency precisely equals
Thus, below the cutoff frequency, phase is little affected, but at higher frequencies, the
phase shift caused by the circuit becomes
This phase shift corresponds to the difference between a cosine and a sine.
We can use the transfer function to find the output when the input voltage is a
sinusoid for two reasons. First of all, a sinusoid is the sum of two complex
exponentials, each having a frequency equal to the negative of the other. Secondly,
because the circuit is linear, superposition applies. If the source is a sine wave, we
know that
Since the input is the sum of two complex exponentials, we know that the output is
also a sum of two similar complex exponentials, the only difference being that the
complex amplitude of each is multiplied by the transfer function evaluated at each
exponential's frequency.
As noted earlier, the transfer function is most conveniently expressed in polar form: H
(f)=|H (f) |e
j∠(H(f)
. Furthermore, |H (−f) | = |H (f) | (even symmetry of the magnitude)
and ∠ (H (−f)) = −(∠ (H (f)) (odd symmetry of the phase). The output voltage expression
simplifes to
76
The circuit's output to a sinusoidal input is also a sinusoid, having a gain equal
to the magnitude of the circuit's transfer function evaluated at the source
frequency and a phase equal to the phase of the transfer function at the source
frequency. It will turn out that this input-output relation description applies to any
linear circuit having a sinusoidal source.
Exercise 3.13.1
This input-output property is a special case of a more general result.
Show that if the source can be written as the imaginary part of a
complex exponential vin (t) = Im (Vej2πft) the output is given byvout (t)
= Im(VH (f) e j2πft). Show that a similar result also holds for the real
part.
The notion of impedance arises when we assume the sources are complex
exponentials. This assumption may seem restrictive; what would we do if the source
were a unit step? When we use impedances to find the transfer function between the
source and the output variable, we can derive from it the differential equation that
relates input and output. The differential equation applies no matter what the source
may be. As we have argued, it is far simpler to use impedances to find the differential
equation (because we can use series and parallel combination rules) than any other
method. In this sense, we have not lost anything by temporarily pretending the source
is a complex exponential.
In fact we can also solve the differential equation using impedances! Thus, despite the
apparent restrictiveness of impedances, assuming complex exponential sources is
actually quite general.
3.14 Designing Transfer Functions
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
If the source consists of two (or more) signals, we know from linear system theory that
the output voltage equals the sum of the outputs produced by each signal alone. In
short, linear circuits are a special case of linear systems, and therefore superposition
applies. In particular, suppose these component signals are complex exponentials,
each of which has a frequency different from the others. The transfer function
portrays how the circuit affects the amplitude and phase of each component, allowing
us to understand how the circuit works on a complicated signal. Those components
having a frequency less than the cutoff frequency pass through the circuit with little
modification while those having higher frequencies are suppressed. The circuit is said
to act as a filter, filtering the source signal based on the frequency of each
component complex exponential. Because low frequencies pass through the filter, we
call it a lowpass filter to express more precisely its function.
We have also found the ease of calculating the output for sinusoidal inputs through
the use of the transfer function. Once we find the transfer function, we can write the
output directly as indicated by the output of a circuit for a sinusoidal input (3.18).
77
Example 3.5
Figure 3.31 RL circuit
Let's apply these results to a final example, in which the input is a
voltage source and the output is the inductor current. The source
voltage equals Vin = 2cos (2π60t)+3. We want the circuit to pass
constant (offset) voltage essentially unaltered (save for the fact that
the output is a current rather than a voltage) and remove the 60 Hz
term. Because the input is the sum of two sinusoids a constant is a
zero-frequency cosine our approach is
1.
2.
3.
4.
find the transfer function using impedances;
use it to find the output due to each input component;
add the results;
find element values that accomplish our design criteria.
Because the circuit is a series combination of elements, let's use voltage divider to find
the transfer function between Vin and V , then use the v-i relation of the inductor to
find its current.
where
and
[Do the units check?] The form of this transfer function should be familiar; it is a
lowpass filter, and it will perform our desired function once we choose element values
properly. The constant term is easiest to handle. The output is given by
. Thus, the value we choose for the resistance will determine the scaling factor of how
voltage is converted into current. For the 60 Hz component signal, the output current
is 2|H (60) |cos (2π60t + ∠ (H (60)). The total output due to our source is
78
The cutoff frequency for this filter occurs when the real and imaginary parts of the
transfer function's denominator equal each other. Thus, 2πfcL = R, which gives
. We want this cutoff frequency to be much less than 60 Hz. Suppose we place it at,
say, 10 Hz. This specification would require the component values to be related by
The transfer function at 60 Hz would be
which yields an attenuation (relative to the gain at zero frequency) of about 1/6, and
result in
an output amplitude of
relative to the constant term's amplitude of
. A factor of 10 relative size between the two components seems reasonable. Having a
100 mH inductor would require a 6.28 Ω resistor. An easily available resistor value is
6.8 Ω; thus, this choice results in cheaply and easily purchased parts. To make the
resistance bigger would require a proportionally larger inductor. Unfortunately, even a
1 H inductor is physically large; consequently low cutoff frequencies require smallvalued resistors and large-valued inductors. The choice made here represents only
one compromise.
The phase of the 60 Hz component will very nearly be
leaving it to be
. The waveforms for the input and output are shown in Figure 3.32.
Waveforms
79
:
Figure 3.32 Waveforms Input and output waveforms for the example RL circuit when the element values
are R =6.28Ω and L = 100mH.
Note that the sinusoid's phase has indeed shifted; the lowpass filter not only reduced
â—¦
the 60 Hz signal's amplitude, but also shifted its phase by 90 .
3.15 Formal Circuit Methods: Node Method
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
In some (complicated) cases, we cannot use the simplification techniques such as
parallel or series combination rules to solve for a circuit's input-output relation. In
other modules, we wrote v-i relations and Kirchhoff’s laws haphazardly, solving them
more on intuition than procedure. We need a formal method that produces a small,
easy set of equations that lead directly to the input-output relation we seek. One such
technique is the node method.
80
Figure 3.33 Node Voltage
The node method begins by finding all nodes places where circuit elements attach to
each other in the circuit. We call one of the nodes the reference node; the choice of
reference node is arbitrary, but it is usually chosen to be a point of symmetry or the
"bottom" node. For the remaining nodes, we define node voltages en that represent
the voltage between the node and the reference. These node voltages constitute the
only unknowns; all we need is a sufficient number of equations to solve for them. In
our example, we have two node voltages. The very act of defining node voltages is
equivalent to using all the KVL equations at your disposal. The reason for this
simple, but astounding, fact is that a node voltage is uniquely defined regardless of
what path is traced between the node and the reference. Because two paths between
a node and reference have the same voltage, the sum of voltages around the loop
equals zero.
In some cases, a node voltage corresponds exactly to the voltage across a voltage
source. In such cases, the node voltage is specifed by the source and is not an
unknown. For example, in our circuit, e1 = vin; thus, we need only to find one node
voltage.
The equations governing the node voltages are obtained by writing KCL equations at
each node having an unknown node voltage, using the v-i relations for each element.
In our example, the only circuit equation is
A little refection reveals that when writing the KCL equations for the sum of currents
leaving a node, that node's voltage will always appear with a plus sign, and all other
node voltages with a minus sign. Systematic application of this procedure makes it
easy to write node equations and to check them before solving them. Also remember
to check units at this point: Every term should have units of current. In our example,
solving for the unknown node voltage is easy:
Have we really solved the circuit with the node method? Along the way, we have used
KVL, KCL, and the v-i relations. Previously, we indicated that the set of equations
resulting from applying these laws is necessary and sufficient. This result guarantees
that the node method can be used to "solve" any circuit. One fallout of this result is
that we must be able to find any circuit variable given the node voltages and sources.
All circuit variables can be found using the v-i relations and voltage divider. For
example, the current through R3 equals
81
.
Figure 3.34 circuit example
The presence of a current source in the circuit does not afect the node method
greatly; just include it in writing KCL equations as a current leaving the node. The
circuit has three nodes, requiring us to define two node voltages. The node equations
are
(Node 1)
(Node 2)
Note that the node voltage corresponding to the node that we are writing KCL for
enters with a positive sign, the others with a negative sign, and that the units of each
term is given in amperes. Rewrite these equations in the standard set-of-linearequations form.
Solving these equations gives
To find the indicated current, we simply use
.
82
Example 3.6: Node Method Example
Figure 3.35 circuit example 2
In this circuit (Figure 3.35), we cannot use the series/parallel
combination rules: The vertical resistor at node 1 keeps the two
horizontal 1 Ω resistors from being in series, and the 2 Ω resistor
prevents the two 1 Ω resistors at node 2 from being in series. We
really do need the node method to solve this circuit! Despite having
six elements, we need only define two node voltages. The node
equations are
(Node 1)
(Node 2)
Solving these equations yields
and
. The output current equals
One unfortunate consequence of using the element's numeric values
from the outset is that it becomes impossible to check units while
setting up and solving equations.
83
Exercise 3.15.1
What is the equivalent resistance seen by the voltage source?
Figure 3.36 Node Method and Impedances Modification of the circuit shown on the left to
illustrate the node method and the effect of adding the resistor R2.
The node method applies to RLC circuits, without significant
modification from the methods used on simple resistive circuits, if
we use complex amplitudes. We rely on the fact that complex
amplitudes satisfy KVL, KCL, and impedance-based v-i relations. In
the example circuit, we define complex amplitudes for the input and
output variables and for the node voltages. We need only one node
voltage here, and its KCL equation is
with the result
To find the transfer function between input and output voltages, we
compute the ratio
. The transfer function's magnitude and angle are
To find the transfer function between input and output voltages, we compute the ratio
. The transfer function's magnitude and angle are
This circuit differs from the one shown previously (Figure 3.29: Simple Circuit) in that
the resistor R2 has been added across the output. What effect has it had on the
transfer function, which in the original circuit was a lowpass filter having cutoff
frequency
84
As shown in Figure 3.37, adding the second resistor has two effects: it lowers the gain
in the passband (the range of frequencies for which the filter has little effect on the
input) and increases the cutoff frequency.
Figure 3.37 Transfer Function Transfer functions of the circuits shown in Figure 3.36 (Node Method and
Impedances). Here, R1 =1, R2 =1, and C = 1.
When R2 = R1, as shown on the plot, the passband gain becomes half of the original,
and the cutoff frequency increases by the same factor. Thus, adding R2 provides a
'knob' by which we can trade passband gain for cutoff frequency.
Exercise 3.15.2
We can change the cutoff frequency without affecting passband gain
by changing the resistance in the original circuit. Does the addition of
the R2 resistor help in circuit design?
3.16 Power Conservation in Circuits
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Now that we have a formal method the node method for solving circuits, we can use it
to prove a powerful result: KVL and KCL are all that are required to show that all
circuits conserve power, regardless of what elements are used to build the circuit.
Figure 3.38 Part of a general circuit to prove Conservation of Power
85
First of all, define node voltages for all nodes in a given circuit. Any node chosen as the
reference will do. For example, in the portion of a large circuit (Figure 3.38) depicted
here, we define node voltages for nodes a, b and c. With these node voltages, we can
express the voltage across any element in terms of them. For example, the voltage
across element 1 is given by vl= eb− ea. The instantaneous power for element 1
becomes
Writing the power for the other elements, we have
When we add together the element power terms, we discover that once we collect
terms involving a particular node voltage, it is multiplied by the sum of currents
leaving the node minus the sum of currents entering. For example, for node b, we
have eb (i3 − il ). We see that the currents will obey KCL that multiply each node
voltage. Consequently, we conclude that the sum of element powers must equal
zero in any circuit regardless of the elements used to construct the circuit.
The simplicity and generality with which we proved this results generalizes to other
situations as well. In particular, note that the complex amplitudes of voltages and
currents obey KVL and KCL, respectively.
Consequently, we have that
Furthermore, the complex-conjugate of currents also satisfies KCL, which means we
also have
. And finally, we know that evaluating the real-part of an expression is linear. Finding
the real-part of this power conservation gives the result that average power is also
conserved in any circuit.
Note: This proof of power conservation can be generalized in another very
interesting way. All we need is a set of voltages that obey KVL and a set of
currents that obey KCL. Thus, for a given circuit topology (the specific way
elements are interconnected), the voltages and currents can be measured at
different times and the sum of v-i products is zero.
86
Even more interesting is the fact that the elements don't matter. We can take a circuit
and measure all the voltages. We can then make element-for-element replacements
and, if the topology has not changed, we can measure a set of currents. The sum of
the product of element voltages and currents will also be zero!
3.17 Electronics
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
So far we have analyzed electrical circuits: The source signal has more power than
the output variable, be it a voltage or a current. Power has not been explicitly defined,
but no matter. Resistors, inductors, and capacitors as individual elements certainly
provide no power gain, and circuits built of them will not magically do so either. Such
circuits are termed electrical in distinction to those that do provide power gain:
electronic circuits. Providing power gain, such as your stereo reading a CD and
producing sound, is accomplished by semiconductor circuits that contain transistors.
The basic idea of the transistor is to let the weak input signal modulate a strong
current provided by a source of electrical power the power supply to produce a more
powerful signal. A physical analogy is a water faucet: By turning the faucet back and
forth, the water flow varies accordingly, and has much more power than expended in
turning the handle. The waterpower results from the static pressure of the water in
your plumbing created by the water utility pumping the water up to your local water
tower. The power supply is like the water tower, and the faucet is the transistor, with
the turning achieved by the input signal. Just as in this analogy, a power supply is a
source of constant voltage as the water tower is supposed to provide a constant water
pressure.
A device that is much more convenient for providing gain (and other useful features as
well) than the transistor is the operational amplifier, also known as the op-amp. An
op-amp is an integrated circuit (a complicated circuit involving several transistors
constructed on a chip) that provides a large voltage gain if you attach the power
supply. We can model the op-amp with a new circuit element: the dependent source.
3.18 Dependent Sources
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A dependent source is either a voltage or current source whose value is proportional
to some other voltage or current in the circuit. Thus, there are four different kinds of
dependent sources; to describe an op-amp, we need a voltage-dependent voltage
source. However, the standard circuit-theoretical model for a transistor contains a
current-dependent current source. Dependent sources do not serve as inputs to a
circuit like independent sources. They are used to model active circuits: those
containing electronic elements. The RLC circuits we have been considering so far are
known as passive circuits.
87
Figure 3.39 dependent sources Of the four possible dependent sources, depicted is a voltage-dependent
voltage source in the context of a generic circuit.
Figure 3.40 shows the circuit symbol for the op-amp and its equivalent circuit in terms
of a voltage-dependent voltage source.
Figure 3.40 op-amp The op-amp has four terminals to which connections can be made. Inputs attach to
nodes a and b, and the output is node c. As the circuit model on the right shows, the op-amp serves as an
amplifier for the difference of the input node voltages.
Here, the output voltage equals an amplified version of the difference of node
voltages appearing across its inputs. The dependent source model portrays how the
op-amp works quite well. As in most active circuit schematics, the power supply is not
shown, but must be present for the circuit model to be accurate. Most operational
amplifiers require both positive and negative supply voltages for proper operation.
Because dependent sources cannot be described as impedances, and because the
dependent variable cannot "disappear" when you apply parallel/series combining
rules, circuit simplifications such as current and voltage divider should not be applied
in most cases. Analysis of circuits containing dependent sources essentially requires
use of formal methods, like the node method (Section 3.15). Using the node method
for such circuits is not difficult, with node voltages defined across the source treated
as if they were known (as with independent sources). Consider the circuit shown on
the top in Figure 3.41.
88
:
Figure 3.41 feedback op-amp The top circuit depicts an op-amp in a feedback amplifier confguration. On
the bottom is the equivalent circuit, and integrates the op-amp circuit model into the circuit.
Note that the op-amp is placed in the circuit "upside-down," with its inverting input at
the top and serving as the only input. As we explore op-amps in more detail in the
next section, this configuration will appear again and again and its usefulness
demonstrated. To determine how the output voltage is related to the input voltage,
we apply the node method. Only two node voltages v and vout need be defined; the
remaining nodes are across sources or serve as the reference. The node equations are
Note that no special considerations were used in applying the node method to this
dependent-source circuit. Solving these to learn how vout relates to vin yields
This expression represents the general input-output relation for this circuit, known as
the standard feedback configuration. Once we learn more about Operational
Amplifiers (Page 89), in particular what its typical element values are, the expression
will simplify greatly. Do note that the units check, and that the parameter G of the
dependent source is a dimensionless gain.
89
3.19 Operational Amplifiers
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 3.42 Op-Amp The op-amp has four terminals to which connections can be made. Inputs attach to
nodes a and b, and the output is node c. As the circuit model on the right shows, the op-amp serves as an
amplifier for the difference of the input node voltages.
Op-amps not only have the circuit model shown in Figure 3.42, but their element
values are very special.
• The input resistance, Rin, is typically large, on the order of 1 MΩ.
• The output resistance, Rout, is small, usually less than 100 Ω.
5
• The voltage gain, G, is large, exceeding 10 .
The large gain catches the eye; it suggests that an op-amp could turn a 1 mV input
signal into a 100 V one. If you were to build such a circuit attaching a voltage source to
node a, attaching node b to the reference, and looking at the output you would be
disappointed. In dealing with electronic components, you cannot forget the
unrepresented but needed power supply.
UNMODELED LIMITATIONS IMPOSED BY POWER SUPPLIES: It is impossible for electronic
components to yield voltages that exceed those provided by the power supply or for
them to yield currents that exceed the power supply's rating.
Typical power supply voltages required for op-amp circuits are ± (15V ). Attaching the 1
mv signal not only would fail to produce a 100 V signal, the resulting waveform would
be severely distorted. While a desirable outcome if you are a rock & roll aficionado,
high-quality stereos should not distort signals. Another consideration in designing
circuits with op-amps is that these element values are typical: Careful control of the
gain can only be obtained by choosing a circuit so that its element values dictate the
resulting gain, which must be smaller than that provided by the op-amp.
90
Figure 3.43 op-amp 2 The top circuit depicts an op-amp in a feedback amplifier configuration. On the
bottom is the equivalent circuit, and integrates the op-amp circuit model into the circuit.
3.20 Inverting Amplifier
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The feedback configuration shown in Figure 3.43 is the most common op-amp circuit
for obtaining what is known as an inverting amplifier.
provides the exact input-output relationship. In choosing element values with respect
to op-amp characteristics, we can simplify the expression dramatically.
• Make the load resistance, RL, much larger than Rout. This situation drops the term
from the second factor of (3.27).
• Make the resistor, R, smaller than Rin, which means that the
term in the third factor is negligible.
With these two design criteria, the expression becomes
Because the gain is large and the resistance Rout is small, the first term becomes
, leaving us with
91
If we select the values of RF and R so that (GR » RF), this factor will no longer depend
on the op-amp's inherent gain, and it will equal
.
Under these conditions, we obtain the classic input-output relationship for the opamp-based inverting amplifier.
Consequently, the gain provided by our circuit is entirely determined by our choice of
the feedback resistor RF and the input resistor R. It is always negative, and can be less
than one or greater than one in magnitude. It cannot exceed the op-amp's inherent
gain and should not produce such large outputs that distortion results (remember the
power supply!). Interestingly, note that this relationship does not depend on the load
resistance. This effect occurs because we use load resistances large compared to the
op-amp's output resistance. Thus observation means that, if careful, we can place opamp circuits in cascade, without incurring the effect of succeeding circuits changing
the behavior (transfer function) of previous ones; see this problem (Problem 3.44).
3.21 Active Filters
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
As long as design requirements are met, the input-output relation for the inverting
amplifier also applies when the feedback and input circuit elements are impedances
(resistors, capacitors, and inductors).
Figure 3.44 op-amp
92
Example 3.7
Let's design an op-amp circuit that functions as a lowpass filter. We
want the transfer function between the output and input voltage to
be
where K equals the passband gain and fc is the cutoff frequency. Let's
assume that the inversion (negative gain) does not matter. With the
transfer function of the above op-amp circuit in mind, let's consider
some choices.
• . This choice means the feedback impedance is a resistor and that the input
impedance is a series combination of an inductor and a resistor. In circuit design,
we try to avoid inductors because they are physically bulkier than capacitors.
• . Consider the reciprocal of the feedback impedance (its admittance):
• . Since this admittance is a sum of admittances, this expression suggests the
parallel combination of a resistor (value 1 Ω) and a capacitor (value
). We have the right idea, but the values (like 1 Ω) are not right. Consider the
general RC parallel combination: its admittance is
. Letting the input resistance equal R, the transfer function of the op-amp
inverting amplifier now is
and the cutoff frequency
Creating a specific transfer function with op-amps does not have a unique answer. As
opposed to design with passive circuits, electronics is more flexible (a cascade of
circuits can be built so that each has little effect on the others; see Problem 3.44) and
gain (increase in power and amplitude) can result. To complete our example, let's
assume we want a lowpass filter that emulates what the telephone companies do.
Signals transmitted over the telephone have an upper frequency limit of about 3 kHz.
93
For the second design choice, we require RF C =3.3×10
−4
. Thus, many choices for
resistance and capacitance values are possible. A 1 µF capacitor and a 330 Ω resistor,
10 nF and 33 kΩ, and 10 pF and 33 MΩ would all theoretically work. Let's also desire a
voltage gain of ten:
which means
Recall that we must have R<Rin. As the op-amp's input impedance is about 1 MΩ, we
don't want R too large, and this requirement means that the last choice for resistor/
capacitor values won't work. We also need to ask for less gain than the op-amp can
provide itself. Because the feedback "element" is an impedance (a parallel resistor
capacitor combination), we need to examine the gain requirement more carefully. We
must have
for all frequencies of interest. Thus,
. As this impedance decreases with frequency, the design specification of
means that this criterion is easily met. Thus, the first two choices for the resistor and
capacitor values (as well as many others in this range) will work well. Additional
considerations like parts cost might enter into the picture. Unless you have a highpower application (this isn't one) or ask for high-precision components, costs don't
depend heavily on component values as long as you stay close to standard values. For
d
resistors, having values r10 , easily obtained values of r are 1, 1.4, 3.3, 4.7, and 6.8,
and the decades span 0-8.
Exercise 3.19.1
What is special about the resistor values; why these rather oddappearing values for r?
3.22 Intuitive Way of Solving Op-Amp Circuits
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we meet op-amp design specifications, we can simplify our circuit calculations
greatly, so much so that we don't need the op-amp's circuit model to determine the
transfer function. Here is our inverting amplifier.
94
Figure 3.45 op-amp
Figure 3.46 op-amp 2
When we take advantage of the op-amp's characteristics large input impedance, large
gain, and small output impedance we note the two following important facts.
• The current iin must be very small. The voltage produced by the dependent
5
source is 10 times the voltage v. Thus, the voltage v must be small, which means
that
−5
in must be tiny. For example, if the output is about 1 V, the voltage v = 10 V,
making the current iin = 10
−11
A. Consequently, we can ignore iin in our
calculations and assume it to be zero.
• Because of this assumption essentially no current flow through Rin the voltage v
must also be essentially zero. This means that in op-amp circuits, the voltage
across the op-amp's input is basically zero.
Armed with these approximations, let's return to our original circuit as shown in
Figure 3.46. The node voltage e is essentially zero, meaning that it is essentially tied to
the reference node. Thus, the current through the resistor R equals
. Furthermore, the feedback resistor appears in parallel with the load resistor.
Because the current going into the op-amp is zero, all of the current flowing through R
flows through the feedback resistor (iF = i)! The voltage across the feedback resistor v
equals
95
. Because the left end of the feedback resistor is essentially attached to the reference
node, the voltage across it equals the negative of that across the output resistor:
. Using this approach makes analyzing new op-amp circuits much easier. When using
this technique, check to make sure the results you obtain are consistent with the
assumptions of essentially zero current entering the op-amp and nearly zero voltage
across the op-amp's inputs.
Example 3.8
Figure 3.47 Two Source Circuit Two-source, single-output op-amp circuit example.
Let's try this analysis technique on a simple extension of the
inverting amplifier confguration shown in Figure 3.47 (Two Source
Circuit). If either of the source-resistor combinations were not
present, the inverting amplifier remains, and we know that transfer
function. By superposition, we know that the input-output relation is
When we start from scratch, the node joining the three resistors is at
the same potential as the reference,
, and the sum of currents flowing into that node is zero. Thus, the
current i flowing in the resistor RF equals
. Because the feedback resistor is essentially in parallel with the load
resistor, the voltages must satisfy v = −vout. In this way, we obtain the
input-output relation given above.
What utility does this circuit have? Can the basic notion of the circuit
be extended without bound?
96
3.23 The Diode
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
:
Figure 3.48 Diode v-i relation and schematic symbol for the diode. Here, the diode parameters were room
temperature and I0 =1 µA.
The resistor, capacitor, and inductor are linear circuit elements in that their v-i
relations are linear in the mathematical sense. Voltage and current sources are
(technically) nonlinear devices: stated simply, doubling the current through a voltage
source does not double the voltage. A more blatant, and very useful, nonlinear circuit
element is the diode (learn more34 ). Its input-output relation has an exponential
form.
Here, the quantity q represents the charge of a single electron in coulombs, k is
Boltzmann's constant, and T is the diode's temperature in K. At room temperature, the
ratio
The constant I0 is the leakage current, and is usually very small. Viewing this v-i
relation in Figure 3.48 (Diode), the nonlinearity becomes obvious. When the voltage is
positive, current flows easily through the diode. This situation is known as forward
biasing. When we apply a negative voltage, the current is quite small, and equals I0,
known as the leakage or reverse-bias current. A less detailed model for the diode
has any positive current flowing through the diode when it is forward biased, and no
current when negative biased. Note that the diode's schematic symbol looks like an
arrowhead; the direction of current flow corresponds to the direction the arrowhead
points.
Figure 3.49 diode circuit
97
Because of the diode's nonlinear nature, we cannot use impedances nor series/
parallel combination rules to analyze circuits containing them. The reliable node
method can always be used; it only relies on KVL for its application, and KVL is a
statement about voltage drops around a closed path regardless of whether the
elements are linear or not. Thus, for this simple circuit we have
This equation cannot be solved in closed form. We must understand what is going on
from basic principles, using computational and graphical aids. As an approximation,
when vin is positive, current flows through the diode so long as the voltage vout is
smaller than vin (so the diode is forward biased). If the source is negative or vout
"tries" to be bigger than vin, the diode is reverse-biased, and the reverse-bias current
flows through the diode. Thus, at this level of analysis, positive input voltages result in
positive output voltages with negative ones resulting in vout = −(RI0).
Figure 3.50 diode circuit
We need to detail the exponential nonlinearity to determine how the circuit distorts
the input voltage waveform. We can of course numerically solve Figure 3.50 to
determine the output voltage when the input is a sinusoid. To learn more, let's
express this equation graphically. We plot each term as a function of vout for various
values of the input voltage vin; where they intersect gives us the output voltage. The
left side, the current through the output resistor, does not vary itself with vin, and
thus we have a fixed straight line. As for the right side, which expresses the diode's v-i
relation, the point at which the curve crosses the vout axis gives us the value of vin.
Clearly, the two curves will always intersect just once for any value of vin, and for
positive vin the intersection occurs at a value for voutsmaller than vin. This reduction
is smaller if the straight line has a shallower slope, which corresponds to using a
bigger output resistor. For negative vin, the diode is reverse-biased and the output
voltage equals − (RI0).
What utility might this simple circuit have? The diode's nonlinearity cannot be escaped
here, and the clearly evident distortion must have some practical application if the
circuit were to be useful. This circuit, known as a half-wave rectifier, is present in
virtually every AM radio twice and each serves very different functions! We'll learn
what functions later.
98
Figure 3.51 diode circuit 3
Here is a circuit involving a diode that is actually simpler to analyze than the previous
one. We know that the current through the resistor must equal that through the
diode. Thus, the diode's current is proportional to the input voltage. As the voltage
across the diode is related to the logarithm of its current, we see that the input-output
relation is
Clearly, the name logarithmic amplifier is justified for this circuit.
3.24 Analog Signal Processing Problems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Problem 3.1: Simple Circuit Analysis
Figure 3.52 Simple Circuit Analysis
For each circuit shown in Figure 3.52, the current i equals cos (2πt).
1. What is the voltage across each element and what is the voltage v
in each case?
2. For the last circuit, are there element values that make the voltage
v equal zero for all time? If so, what element values work?
3. Again, for the last circuit, if zero voltage were possible, what
circuit element could substitute for the capacitor-inductor series
combination that would yield the same voltage?
99
Problem 3.2: Solving Simple Circuits
1. Write the set of equations that govern Circuit A's (Figure 3.53)
behavior.
2. Solve these equations for i1: In other words, express this current in
terms of element and source values by eliminating non-source
voltages and currents.
3. For Circuit B, find the value for RL that results in a current of 5 A
passing through it.
4. What is the power dissipated by the load resistor RL in this case?
Figure 3.53 Solving Simple Circuits
Problem 3.3: Equivalent Resistance
For each of the following circuits (Figure 3.54), find the equivalent
resistance using series and parallel combination rules.
Figure 3.54 Equivalent Resistance
Calculate the conductance seen at the terminals for circuit (c) in
terms of each element's conductance. Compare this equivalent
conductance formula with the equivalent resistance formula you
found for circuit (b). How is the circuit (c) derived from circuit (b)?
100
Problem 3.4: Superposition Principle
One of the most important consequences of circuit laws is the
Superposition Principle: The current or voltage defined for any
element equals the sum of the currents or voltages produced in the
element by the independent sources. This Principle has important
consequences in simplifying the calculation of circuit variables in
multiple source circuits.
Figure 3.55 Superposition Principle
1. For the depicted circuit (Figure 3.55), find the indicated current
using any technique you like (you should use the simplest).
2. You should have found that the current i is a linear combination of
the two source values:i = C1vin +C2iin. This result means that we
can think of the current as a superposition of two components,
each of which is due to a source. We can find each component by
setting the other sources to zero. Thus, to find the voltage source
component, you can set the current source to zero (an open
circuit) and use the usual tricks. To find the current source
component, you would set the voltage source to zero (a short
circuit) and find the resulting current. Calculate the total current i
using the Superposition Principle. Is applying the Superposition
Principle easier than the technique you used in part (1)?
101
Problem 3.5: Current and Voltage Divider
Use current or voltage divider rules to calculate the indicated circuit
variables in Figure 3.56.
Figure 3.56 Current and Voltage Divider
Problem 3.6: Thevenin and MayerNorton Equivalents
Find the Thevenin and Mayer-Norton equivalent circuits for the
following circuits (Figure 3.57).
Figure 3.57 Thevenin and Mayer-Norton Equivalents
102
Problem 3.7: Detective Work
In the depicted circuit (Figure 3.58), the circuit N1 has the v-i relation
v1 =3i1 +7 when is =2.
1. Find the Thevenin equivalent circuit for circuit N2.
2. With is =2, determine R such that i1 = −1.
Figure 3.58 Detective Work
Problem 3.8: Bridge Circuits
Circuits having the form of Figure 3.59 are termed bridge circuits.
Figure 3.59 Bridge Circuits
1. What resistance does the current source see when nothing is
connected to the output terminals?
2. What resistor values, if any, will result in a zero voltage for vout?
3. Assume R1 = 1Ω, R2 = 2Ω, R3 = 2Ω and R4 = 4Ω. Find the current i
when the current source iin is Im ((4+2j) ej2π20t . Express your
answer as a sinusoid.
103
Problem 3.9: Cartesian to Polar
Conversion
Convert the following expressions into polar form. Plot their location
in the complex plane.
Problem 3.10: The Complex Plane
The complex variable z is related to the real variable u according to
• Sketch the contour of values z takes on in the complex plane.
• What are the maximum and minimum values attainable by |z|?
• Sketch the contour the rational function
traces in the complex plane.
Problem 3.11: Cool Curves
In the following expressions, the variable x runs from zero to infinity.
What geometric shapes do the following trace in the complex plane?
104
Problem 3.12: Trigonometric Identities
and Complex Exponentials
Show the following trigonometric identities using complex
exponentials. In many cases, they were derived using this approach.
Figure 3.61Problem 3.13: Transfer
Functions
Find the transfer function relating the complex amplitudes of the
indicated variable and the source shown in Figure 3.60. Plot the
magnitude and phase of the transfer function.
Figure 3.60 Transfer Function
105
Problem 3.14: Using Impedances
Find the differential equation relating the indicated variable to the
source(s) using impedances for each circuit shown in
Figure 3.61 Using Impedances
Problem 3.15: Measurement Chaos
The following simple circuit (Figure 3.62) was constructed but the
signal measurements were made haphazardly. When the source was
sin (2πf0t), the current i (t) equaled
and the voltage
Figure 3.62 Measurement Chaos
1. What is the voltage v1 (t)?
2. Find the impedances Z1 and Z2.
3. Construct these impedances from elementary circuit elements.
106
Problem 3.16: Transfer Functions
In the following circuit (Figure 3.63), the voltage source equals
Figure 3.63 Transfer Functions
1. Find the transfer function between the source and the indicated
output voltage.
2. For the given source, find the output voltage.
Problem 3.17: A Simple Circuit
You are given this simple circuit (Figure 3.64).
Figure 3.64 A Simple Circuit
1. What is the transfer function between the source and the indicated
output current?
2. If the output current is measured to be cos (2t), what was the
source?
107
Problem 3.18: Circuit Design
Figure 3.65 Circuit Design
1. Find the transfer function between the input and the output
voltages for the circuits shown in Figure 3.65.
2. At what frequency does the transfer function have a phase shift of
zero? What is the circuit's gain at this frequency?
3. Specifications demand that this circuit have an output impedance
(its equivalent impedance) less than 8Ω for frequencies above 1
kHz, the frequency at which the transfer function is maximum.
Find element values that satisfy this criterion.
Problem 3.19: Equivalent Circuits and
Power
Suppose we have an arbitrary circuit of resistors that we collapse into
an equivalent resistor using the series and parallel rules. Is the power
dissipated by the equivalent resistor equal to the sum of the powers
dissipated by the actual resistors comprising the circuit? Let's start
with simple cases and build up to a complete proof.
1. Suppose resistors R1 and R2 are connected in parallel. Show that
the power dissipated by R1 I R1 equals the sum of the powers
dissipated by the component resistors.
2. Now suppose R1 and R2 are connected in series. Show the same
result for this combination.
3. Use these two results to prove the general result we seek.
108
Problem 3.20: Power Transmission
The network shown in the figure represents a simple power
transmission system. The generator produces 60 Hz and is modeled
by a simple Thevenin equivalent. The transmission line consists of a
long length of copper wire and can be accurately described as a 50Ω
resistor.
1. Determine the load current RL and the average power the
generator must produce so that the load receives 1,000 watts of
average power. Why does the generator need to generate more
than 1,000 watts of average power to meet this requirement?
2. Suppose the load is changed to that shown in the second figure.
Now how much power must the generator produce to meet the
same power requirement? Why is it more than it had to produce to
meet the requirement for the resistive load?
3. The load can be compensated to have a unity power factor (see
exercise (Exercise 3.11.2)) so that the voltage and current are in
phase for maximum power efficiency. The compensation
technique is to place a circuit in parallel to the load circuit. What
element works and what is its value?
4. With this compensated circuit, how much power must the
generator produce to deliver 1,000 average power to the load?
Figure 3.66 Power Transmission
109
Problem 3.21: Optimal Power
Transmission
The following figure (ds) shows a general model for power
transmission. The power generator is represented by a Thevinin
equivalent and the load by a simple impedance. In most applications,
the source components are fixed while there is some latitude in
choosing the load.
1. Suppose we wanted the maximize "voltage transmission:" make
the voltage across the load as large as possible. What choice of
load impedance creates the largest load voltage? What is the
largest load voltage?
2. If we wanted the maximum current to pass through the load, what
would we choose the load impedance to be? What is this largest
current?
3. What choice for the load impedance maximizes the average power
dissipated in the load? What is most power the generator can
deliver?
Note: One way to maximize a function of a complex variable is to write
the expression in terms of the variable's real and imaginary parts,
evaluate derivatives with respect to each, set both derivatives to zero
and solve the two equations simultaneously.
Figure 3.67 Optimal Power Transmission
110
Problem 3.22: Big is Beautiful
Sammy wants to choose speakers that produce very loud music. He
has an amplifier and notices that the speaker terminals are labeled
"8Ω source."
1. What does this mean in terms of the amplifier’s equivalent
circuit?
2. Any speaker Sammy attaches to the terminals can be wellmodeled as a resistor. Choosing a speaker amounts to choosing
the values for the resistor. What choice would maximize the
voltage across the speakers?
3. Sammy decides that maximizing the power delivered to the
speaker might be a better choice. What values for the speaker
resistor should be chosen to maximize the power delivered to the
speaker?
Problem 3.23: Sharing a Channel
Two transmitter-receiver pairs want to share the same digital
communications channel. The transmitter signals will be added
together by the channel. Receiver design is greatly simplified if first
we remove the unwanted transmission (as much as possible). Each
transmitter signal has the form
where the amplitude is either zero or A and each transmitter uses its
own frequency fi. Each frequency is harmonically related to the bit
interval duration T, where the transmitter 1 uses the frequency
. The datarate is 10Mbps.
1. Draw a block diagram that expresses this communication
scenario.
2. Find circuits that the receivers could employ to separate unwanted
transmissions. Assume the received signal is a voltage and the
output is to be a voltage as well.
3. Find the second transmitter's frequency so that the receivers can
suppress the unwanted transmission by at least a factor of ten.
111
Problem 3.24: Circuit Detective Work
In the lab, the open-circuit voltage measured across an unknown
circuit's terminals equals sin (t). When a 1Ω resistor is place across
the terminals, a voltage of
1. What is the Thevenin equivalent circuit?
2. What voltage will appear if we place a 1 F capacitor across the
terminals?
Problem 3.25: Mystery Circuit
We want to determine as much as we can about the circuit lurking in
the impenetrable box shown in Figure 3.68. A voltage source vin =2 V
has been attached to the left-hand terminals, leaving the right
terminals for tests and measurements.
Figure 3.68 Mystery Circuit
1. Sammy measures v = 10 V when a 1 Ω resistor is attached to the
terminals. Samantha says he is wrong. Who is correct and why?
2. When nothing is attached to the right-hand terminals, a voltage
of v =1 V is measured. What circuit could produce this output?
3. When a current source is attached so that i =2amp, the voltage v is
now 3 V. What resistor circuit would be consistent with this and
the previous part?
112
Problem 3.26: More Circuit Detective
Work
The left terminal pair of a two terminal-pair circuit is attached to a
testing circuit. The test source vin (t) equals sin (t) (Figure 3.69).
Figure 3.69 More Circuit Detective Work
We make the following measurements.
• With nothing attached to the terminals on the right, the voltage v
(t) equals
.
• When a wire is placed across the terminals on the right, the
current i(t) was −(sin(t)).
1. What is the impedance "seen" from the terminals on the right?
2. Find the voltage v (t) if a current source is attached to the
terminals on the right so that i(t) = sin(t).
113
Problem 3.27: Linear, Time-Invariant
Systems
For a system to be completely characterized by a transfer function, it
needs not only be linear, but also to be time-invariant. A system is
said to be time-invariant if delaying the input delays the output by
the same amount. Mathematically, if S (x (t)) = y (t), meaning y (t) is
the output of a systemS (•) when x (t) is the input, S(•) is the timeinvariant if S (x (t − τ)) = y (t − τ) for all delays τ and all inputs x (t).
Note that both linear and nonlinear systems have this property. For
example, a system that squares its input is time-invariant.
1. Show that if a circuit has fixed circuit elements (their values don't
change over time), its input-output relationship is timeinvariant. Hint: Consider the differential equation that describes a
circuit's input-output relationship. What is its general form?
Examine the derivative(s) of delayed signals.
2. Show that impedances cannot characterize time-varying circuit
elements (R, L, and C). Consequently, show that linear, timevarying systems do not have a transfer function.
3. Determine the linearity and time-invariance of the following. Find
the transfer function of the linear, time-invariant (LTI) one(s).
1. diode
2. y (t)= x (t) sin (2πf0t)
3. y (t)= x (t − τ0)
4. y (t)= x (t)+ N (t)
114
Problem 3.28: Long and Sleepless Nights
Sammy went to lab after a long, sleepless night, and constructed the
circuit shown in Figure 3.70. He cannot remember what the circuit,
represented by the impedance Z, was. Clearly, this forgotten circuit is
important as the output is the current passing through it.
1. What is the Thevenin equivalent circuit seen by the impedance?
2. In searching his notes, Sammy fnds that the circuit is to realize
the transfer function
Find the impedance Z as well as values for the other circuit elements.
Figure 3.70 Long and Sleepless Nights
Problem 3.29: A Testing Circuit
The simple circuit here (Figure 3.71) was given on a test.
Figure 3.71 A Testing Circuit
When the voltage source is
the current
1. What is voltage vout (t)?
2. What is the impedance Z at the frequency of the source?
115
Problem 3.30: Black-Box Circuit
You are given a circuit (Figure 3.72) that has two terminals for
attaching circuit elements.
Figure 3.72 Black-Box Circuit
When you attach a voltage source equaling sin (t) to the terminals,
the current through the source equals
When no source is attached (open-circuited terminals), the voltage
across the terminals has the form Asin (4t + φ).
1. What will the terminal current be when you replace the source by a
short circuit?
2. If you were to build a circuit that was identical (from the viewpoint
of the terminals) to the given one, what would your circuit be?
3. For your circuit, what are A and φ?
Problem 3.31: Solving a Mystery Circuit
Sammy must determine as much as he can about a mystery circuit by
attaching elements to the terminal and measuring the resulting
voltage. When he attaches a 1Ω resistor to the circuit's terminals, he
measures the voltage across the terminals to be 3sin (t). When he
attaches a 1F capacitor across the terminals, the voltage is now
1. What voltage should he measure when he attaches nothing to the
mystery circuit?
2. What voltage should Sammy measure if he doubled the size of the
capacitor to 2 F and attached it to the circuit?
116
Problem 3.32: Find the Load Impedance
The depicted circuit (Figure 3.73) has a transfer function between the
output voltage and the source equal to
Figure 3.73 Find the Load Impedance
1. Sketch the magnitude and phase of the transfer function.
2. At what frequency does the phase equal π 2 ?
3. Find a circuit that corresponds to this load impedance. Is your
answer unique? If so, show it to be so; if not, give another
example.
Problem 3.33: Analog "Hum" Rejection
"Hum" refers to corruption from wall socket power that frequently
sneaks into circuits. "Hum" gets its name because it sounds like a
persistent humming sound. We want to find a circuit that will
remove hum from any signal. A Rice engineer suggests using a simple
voltage divider circuit (Figure 3.74) consisting of two series
impedances.
Figure 3.74 Analog "Hum" Rejection 1
1. The impedance Z1 is a resistor. The Rice engineer must decide
between two circuits (Figure 3.75) for the impedance Z2. Which of
these will work?
2. b) Picking one circuit that works, choose circuit element values
that will remove hum.
3. c) Sketch the magnitude of the resulting frequency response.
Figure 3.75 Analog "Hum" Rejection 2
117
Problem3.34: An Interesting Circuit
Figure 3.76 An Interesting Circuit
1. For the circuit shown in Figure 3.76, find the transfer function.
2. What is the output voltage when the input has the form iin = 5sin
(2000πt)?
Problem 3.35: A Simple Circuit
You are given the depicted circuit (Figure 3.77).
Figure 3.77 A Simple Circuit
1. What is the transfer function between the source and the output
voltage?
2. What will the voltage be when the source equals sin (t)?
3. Many function generators produce a constant offset in addition to
a sinusoid. If the source equals 1 + sin(t), what is the output
voltage?
118
Problem 3.36: An Interesting and Useful
Circuit
The depicted circuit (Figure 3.78) has interesting properties, which
are exploited in high-performance oscilloscopes.
Figure 3.78 An Interesting and Useful Circuit
The portion of the circuit labeled "Oscilloscope" represents the scope's input
impedance. R2 = 1MΩ and C2 = 30pF (note the label under the channel 1 input in the
lab's oscilloscopes). A probe is a device to attach an oscilloscope to a circuit, and it has
the indicated circuit inside it.
1. Suppose for a moment that the probe is merely a wire and that the oscilloscope is
attached to a circuit that has a resistive Thevenin equivalent impedance. What
would be the effect of the oscilloscope's input impedance on measured voltages?
2. Using the node method, find the transfer function relating the indicated voltage
to the source when the probe is used.
3. Plot the magnitude and phase of this transfer function when R1 = 9MΩ and C1 =
2pF.
4. For a particular relationship among the element values, the transfer function is
quite simple. Find that relationship and describe what is so special about it.
5. The arrow through C1 indicates that its value can be varied. Select the value for
this capacitor to make the special relationship valid. What is the impedance seen
by the circuit being measured for this special value?
119
Problem 3.37: A Circuit Problem
You are given the depicted circuit (Figure 3.79).
Figure 3.79 A Circuit Problem
1. Find the differential equation relating the output voltage to the
source.
2. What is the impedance "seen" by the capacitor?
Problem 3.38: Analog Computers
Because the differential equations arising in circuits resemble those
that describe mechanical motion, we can use circuit models to
describe mechanical systems. An ELEC 241 student wants to
understand the suspension system on his car. Without a suspension,
the car's body moves in concert with the bumps in the raod. A welldesigned suspension system will smooth out bumpy roads, reducing
the car's vertical motion. If the bumps are very gradual (think of a
hill as a large but very gradual bump), the car's vertical motion
should follow that of the road. The student wants to find a simple
circuit that will model the car's motion. He is trying to decide
between two circuit models (Figure 3.80).
Figure 3.80 Analog Computers
Here, road and car displacements are represented by the voltages
Vroad (t) and Vcar (t), respectively.
1. Which circuit would you pick? Why?
2. For the circuit you picked, what will be the amplitude of the car's
motion if the road has a displacement given by Vroad (t) =
1+sin(2t)?
120
Problem 3.39: Transfer Functions and
Circuits
You are given the depicted network (Figure 3.81).
Figure 3.81 Transfer Functions and Circuits
1. Find the transfer function between Vin and Vout.
2. Sketch the magnitude and phase of your transfer function. Label
important frequency, amplitude and phase values.
3. Find vout (t) when
Problem 3.40: Fun in the Lab
You are given an unopenable box that has two terminals sticking out.
You assume the box contains a circuit. You measure the voltage
across the terminals when nothing is connected to them and the
current
when you place a wire across the terminals.
Problem 3.41: Dependent Sources
Find the voltage Vout in each of the depicted circuits (Figure 3.82).
Figure 3.82 Dependent Sources
121
Problem 3.42: Operational Amplifers
Find the transfer function between the source voltage(s) and the
indicated output voltage for the circuits shown in Figure 3.83.
Figure 3.83 Operational Amplifers
122
Problem 3.43: Op-Amp Circuit
The following circuit (Figure 3.84) is claimed to serve a useful
purpose.
Figure 3.84 Op-Amp Circuit
1. What is the transfer function relating the complex amplitude of
the output signal, the current Iout, to the complex amplitude of
the input, the voltage Vin?
2. What equivalent circuit does the load resistor RL see?
3. Find the output current when
.
123
Problem 3.44: Why Op-Amps are Useful
The circuit (Figure 3.85) of a cascade of op-amp circuits illustrate the
reason why op-amp realizations of transfer functions are so useful.
Figure 3.85 Why Op-Amps are Useful
1. Find the transfer function relating the complex amplitude of the
voltage Vout (t) to the source. Show that this transfer function
equals the product of each stage's transfer function.
2. What is the load impedance appearing across the first op-amp's
output?
3. Figure 3.84 illustrates that sometimes "designs" can go wrong.
Find the transfer function for this op-amp circuit (Figure 3.86),
and then show that it can't work! Why can't it?
Figure 3.86 op-amp circuit
124
Problem 3.45: Operational Amplifiers
Consider the depicted circuit (Figure 3.87).
Figure 3.87 Operational Amplifiers
1. Find the transfer function relating the voltage Vout (t) to the
source.
2. In particular, R1 = 530Ω, C1 =1µF, R2 =5.3kΩ, C2 =0.01µF, and R3 = R4
=5.3kΩ. Characterize the resulting transfer function and determine
what use this circuit might have.
Problem 3.46: Designing a Bandpass
Filter
We want to design a bandpass filter that has transfer the function
Here, fl is the cutoff frequency of the low-frequency edge of the
passband and fh is the cutoff frequency of the high-frequency edge.
We want fl = 1kH z and fh = 10kHz.
1. Plot the magnitude and phase of this frequency response. Label
important amplitude and phase values and the frequencies at
which they occur.
2. Design a bandpass filter that meets these specifications. Specify
component values.
125
Problem 3.47: Pre-emphasis or Deemphasis?
In audio applications, prior to analog-to-digital conversion signals
are passed through what is known as a pre-emphasis circuit that
leaves the low frequencies alone but provides increasing gain at
increasingly higher frequencies beyond some frequency f0. Deemphasis circuits do the opposite and are applied after digital-toanalog conversion. After pre-emphasis, digitization, conversion back
to analog and de-emphasis, the signal's spectrum should be what it
was.
The op-amp circuit here (Figure 3.88) has been designed for preemphasis or de-emphasis (Samantha can't recall which).
Figure 3.88 Pre-emphasis or De-emphasis?
1. Is this a pre-emphasis or de-emphasis circuit? Find the frequency
f0 that defines the transition from low to high frequencies.
2. What is the circuit's output when the input voltage is sin (2πft),
with f = 4kHz?
3. What circuit could perform the opposite function to your answer
for the first part?
126
Problem 3.48: Active Filter
Find the transfer function of the depicted active filter (Figure 3.89).
Figure 3.89 Active Filter
Problem 3.49: This is a filter?
You are given a circuit (Figure 3.90).
Figure 3.90 This is a filter?
1. What is this circuit's transfer function? Plot the magnitude and
phase.
2. If the input signal is the sinusoid sin (2πf0t), what will the output
be when f0 is larger than the filter’s "cutoff frequency"?
127
Problem 3.50: Optical Receivers
In your optical telephone, the receiver circuit had the form shown
(Figure 3.91).
Figure 3.91 Optical Receivers
This circuit served as a transducer, converting light energy into a
voltage vout. The photodiode acts as a current source, producing a
current proportional to the light intensity falling upon it. As is often
the case in this crucial stage, the signals are small and noise can be a
problem. Thus, the op-amp stage serves to boost the signal and to
filter out-of-band noise.
1. Find the transfer function relating light intensity to vout.
2. What should the circuit realizing the feedback impedance Zf be so
that the transducer acts as a 5 kHz lowpass filter?
3. A clever engineer suggests an alternative circuit (Figure 3.92) to
accomplish the same task. Determine whether the idea works or
not. If it does, find the impedance Zin that accomplishes the
lowpass filtering task. If not, show why it does not work.
Figure 3.92 an alternative circuit
128
Problem 3.51: Reverse Engineering
The depicted circuit (Figure 3.93) has been developed by the TBBG
Electronics design group. They are trying to keep its use secret; we,
representing RU Electronics, have discovered the schematic and want
to figure out the intended application. Assume the diode is ideal.
Figure 3.93 Reverse Engineering
1. Assuming the diode is a short-circuit (it has been removed from
the circuit), what is the circuit's
2. transfer function?
3. With the diode in place, what is the circuit's output when the
input voltage is sin (2πf0t)?
4. What function might this circuit have?
3.25 Solutions to Exercises in Chapter 3
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 3.1.1
One kilowatt-hour equals 3,600,000 watt-seconds, which indeed directly corresponds
to 3,600,000 joules.
Solution to Exercise 3.4.1
KCL says that the sum of currents entering or leaving a node must be zero. If we
consider two nodes together as a "supernode", KCL applies as well to currents
entering the combination. Since no currents enter an entire circuit, the sum of
currents must be zero. If we had a two-node circuit, the KCL equation of one must be
the negative of the other, We can combine all but one node in a circuit into a
supernode; KCL for the supernode must be the negative of the remaining node's KCL
equation. Consequently, specifying n − 1 KCL equations always specifies the remaining
one.
Solution to Exercise 3.4.2
The circuit serves as an amplifier having a gain of
129
Solution to Exercise 3.5.1
The power consumed by the resistor R1 can be expressed as
Solution to Exercise 3.5.2
Solution to Exercise 3.6.1
Replacing the current source by a voltage source does not change the fact that the
voltages are identical.
Consequently,
or
. This result does not depend on the resistor R1, which means that we simply have a
resistor (R2) across a voltage source. The two-resistor circuit has no apparent use.
Solution to Exercise 3.6.2
. Thus, a 10% change means that the ratio
must be less than 0.1. A 1% change means that
Solution to Exercise 3.6.3
In a series combination of resistors, the current is the same in each; in a parallel
combination, the voltage is the same. For a series combination, the equivalent
resistance is the sum of the resistances, which will be larger than any component
resistor's value; for a parallel combination, the equivalent conductance is the sum of
the component conductances, which is larger than any component conductance. The
equivalent resistance is therefore smaller than any component resistance.
Solution to Exercise 3.7.1
and
130
(resistor R2 is shorted out in this case) . Thus,
and
Solution to Exercise 3.7.2
and
.
Solution to Exercise 3.10.1
Division by j2πf arises from integrating a complex exponential. Consequently,
Solution to Exercise 3.11.1
For maximum power dissipation, the imaginary part of complex power should be zero.
As the complex power is given by VI * = |V ||I|e
j(φ−θ)
, zero imaginary part occurs
when the phases of the voltage and currents agree.
Solution to Exercise 3.11.2
Pave = Vrms Irmscos (φ − θ). The cosine term is known as the power factor.
Solution to Exercise 3.13.1
The key notion is writing the imaginary part as the difference between a complex
exponential and its complex conjugate:
The response to Vej2
(
(−f) e- j2
πft)
πft
is VH (f) ej2
πft
(
, which means the response to V* e- j2
πft)
is V*H
.
As H(-f) = (H (f)*), the Superposition Principle says that the output to the imaginary part
is Im (VH (f) ej2
(f) ej2
πft
πft
). The same argument holds for the real part: Re(V ej2
πft
) → Re (VH
).
Solution to Exercise 3.15.1
To find the equivalent resistance, we need to find the current flowing through the
voltage source. This current equals the current we have just found plus the current
flowing through the other vertical 1 Ω resistor. This current equals
131
making the total current through the voltage source (flowing out of it)
.
Thus, the equivalent resistance is
Solution to Exercise 3.15.2
Not necessarily, especially if we desire individual knobs for adjusting the gain and the
cutoff frequency.
Solution to Exercise 3.19.1
The ratio between adjacent values is about
132
Chapter 4 Frequency Domain
4.1 Introduction to the Frequency Domain
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
In developing ways of analyzing linear circuits, we invented the impedance method
because it made solving circuits easier. Along the way, we developed the notion of a
circuit's frequency response or transfer function. This notion, which also applies to all
linear, time-invariant systems, describes how the circuit responds to a sinusoidal input
when we express it in terms of a complex exponential. We also learned the
Superposition Principle for linear systems: The system's output to an input consisting
of a sum of two signals is the sum of the system's outputs to each individual
component.
The study of the frequency domain combines these two notions a system's sinusoidal
response is easy to find and a linear system's output to a sum of inputs is the sum of
the individual outputs to develop the crucial idea of a signal's spectrum. We begin by
finding that those signals that can be represented as a sum of sinusoids is very large.
In fact, all signals can be expressed as a superposition of sinusoids.
As this story unfolds, we'll see that information systems rely heavily on spectral ideas.
For example, radio, television, and cellular telephones transmit over different portions
of the spectrum. In fact, spectrum is so important that communications systems are
regulated as to which portions of the spectrum they can use by the Federal
Communications Commission in the United States and by International Treaty for the
world (see Frequency Allocations (Section 7.3)). Calculating the spectrum is easy: The
Fourier transform defines how we can find a signal's spectrum.
4.2 Complex Fourier Series
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
In an earlier module (Exercise 2.3.1), we showed that a square wave could be
expressed as a superposition of pulses. As useful as this decomposition was in this
example, it does not generalize well to other periodic signals: How can a superposition
of pulses equal a smooth signal like a sinusoid? Because of the importance of
sinusoids to linear systems, you might wonder whether they could be added together
to represent a large number of periodic signals. You would be right and in good
company as well. Euler and Gauss in particular worried about this problem, and Jean
Baptiste Fourier5 got the credit even though tough mathematical issues were not
settled until later. They worked on what is now known as the Fourier series:
representing any periodic signal as a superposition of sinusoids.
But the Fourier series goes well beyond being another signal decomposition method.
Rather, the Fourier series begins our journey to appreciate how a signal can be
133
described in either the time-domain or the frequency-domain with no compromise.
Let s (t) be a periodic signal with period T. We want to show that periodic signals,
even those that have constant-valued segments like a square wave, can be expressed
as sum of harmonically related sine waves: sinusoids having frequencies that are
integer multiples of the fundamental frequency. Because the signal has period T ,
the fundamental frequency is
. The complex Fourier series expresses the signal as a superposition of complex
exponentials having frequencies
,
k = {. . ., −1, 0, 1,...}.
with
The real and imaginary parts of the Fourier coefficients ck are written in this unusual
way for convenience in defining the classic Fourier series. The zeroth coefficient
equals the signal's average value and is real-valued for real-valued signals: c0 = a0. The
family of functions
are called basis functions and form the foundation of the Fourier series. No matter
what the periodic signal might be, these functions are always present and form the
representation's building blocks. They depend on the signal period T, and are indexed
by k.
KEY POINT: Assuming we know the period, knowing the Fourier coefficients is
equivalent to knowing the signal. Thus, it makes no difference if we have a timedomain or a frequency-domain characterization of the signal.
Exercise 4.2.1
What is the complex Fourier series for a sinusoid? To find the Fourier
coefficients, we note the orthogonality property
Assuming for the moment that the complex Fourier series "works," we can find a
signal's complex Fourier coefficients, its spectrum, by exploiting the orthogonality
properties of harmonically related complex exponentials. Simply multiply each side of
(4.1) by e
−(j2πlt)
and integrate over the interval [0,T ].
134
Example 4.1
Finding the Fourier series coefficients for the square wave sqT(t) is
very simple. Mathematically, this signal can be expressed as
The expression for the Fourier coefficients has the form
Note: When integrating an expression containing j, treat it just like
any other constant.
The two integrals are very similar, one equaling the negative of the other. The fnal
expression becomes
Thus, the complex Fourier series for the square wave is
Consequently, the square wave equals a sum of complex exponentials, but only those
having frequencies equal to odd multiples of the fundamental frequency
. The coefficients decay slowly as the frequency index k increases. This index
corresponds to the k-th harmonic of the signal's period.
A signal's Fourier series spectrum ck has interesting properties.
Property 4.1:
If s (t) is real, ck = c−k* (real-valued periodic signals have conjugate-symmetric spectra).
This result follows from the integral that calculates the ck from the signal.
Furthermore, this result means that Re (ck)= Re (c−k): The real part of the Fourier
coefficients for real-valued signals is even. Similarly, Im (ck)= −Im (c−k): The imaginary
parts of the Fourier coefficients have odd symmetry. Consequently, if you are given
135
the Fourier coefficients for positive indices and zero and are told the signal is realvalued, you can find the negative-indexed coefficients, hence the entire spectrum. This
kind of symmetry, ck = c−k*, is known as conjugate symmetry.
Property 4.2:
If s (−t)= s (t), which says the signal has even symmetry about the origin, c−k = ck. Given
the previous property for real-valued signals, the Fourier coefficients of even signals
are real-valued. A real-valued Fourier expansion amounts to an expansion in terms of
only cosines, which is the simplest example of an even signal.
Property 4.3:
If s(−t)= −(s(t)), which says the signal has odd symmetry, c−k = −ck. Therefore, the
Fourier coefficients are purely imaginary. The square wave is a great example of an
odd-symmetric signal.
Property 4.4:
The spectral coefficients for a periodic signal delayed by τ, s (t − τ), are
, where ck denotes the spectrum of s(t). Delaying a signal by τ seconds results in a
spectrum having a linear phaseshift of
in comparison to the spectrum of the undelayed signal. Note that the spectral
magnitude is unaffected. Showing this property is easy.
Proof:
Note that the range of integration extends over a period of the integrand.
Consequently, it should not matter how we integrate over a period, which means that
, and we have our result.
The complex Fourier series obeys Parseval's Theorem, one of the most important
results in signal analysis. This general mathematical result says you can calculate a
signal's power in either the time domain or the frequency domain.
Theorem 4.1: Parseval's Theorem
Average power calculated in the time domain equals the power calculated in the
frequency domain.
136
This result is a (simpler) re-expression of how to calculate a signal's power than with
the real-valued Fourier series expression for power. Let's calculate the Fourier
coefficients of the periodic pulse signal shown here (Figure 4.1).
Figure 4.1 Periodic pulse signal
The pulse width is Δ, the period T , and the amplitude A. The complex Fourier
spectrum of this signal is given by
At this point, simplifying this expression requires knowing an interesting property.
Armed with this result, we can simply express the Fourier series coefficients for our
pulse sequence.
Because this signal is real-valued, we find that the coefficients do indeed have
conjugate symmetry: ck= c−k * . The periodic pulse signal has neither even nor odd
symmetry; consequently, no additional symmetry exists in the spectrum. Because the
spectrum is complex valued, to plot it we need to calculate its magnitude and phase.
The function neg (·) equals -1 if its argument is negative and zero otherwise. The
somewhat complicated expression for the phase results because the sine term can be
negative; magnitudes must be positive, leaving the occasional negative values to be
accounted for as a phase shift of π.
137
Figure 4.2 Periodic Pulse Sequence The magnitude and phase of the periodic pulse sequence's spectrum
is shown for positive frequency indices. Here =0.2 and A =1.
Also note the presence of a linear phase term (the first term in ∠ (ck) is proportional to
frequency
). Comparing this term with that predicted from delaying a signal, a delay of
is present in our signal. Advancing the signal by this amount centers the pulse about
the origin, leaving an even signal, which in turn means that its spectrum is real-valued.
Thus, our calculated spectrum is consistent with the properties of the Fourier
spectrum.
Exercise 4.2.2
What is the value of c0? Recalling that this spectral coefficient
corresponds to the signal's average value, does your answer make
sense?
The phase plot shown in Figure 4.2 requires some explanation as it does not seem to
agree with what (4.10) suggests. There, the phase has a linear component, with a jump
of π every time the sinusoidal term changes sign. We must realize that any integer
multiple of 2π can be added to a phase at each frequency without affecting the value
of the complex spectrum. We see that at frequency index 4 the phase is nearly −π. The
phase at index 5 is undefined because the magnitude is zero in this example. At index
6, the formula suggests that the phase of the linear term should be less than −π (more
negative). In addition, we expect a shift of −π in the phase between indices 4 and 6.
Thus, the phase value predicted by the formula is a little less than − (2π). Because we
138
can add 2π without affecting the value of the spectrum at index 6, the result is a
slightly negative number as shown. Thus, the formula and the plot do agree. In phase
calculations like those made in MATLAB, values are usually confined to the range [−π,
π) by adding some (possibly negative) multiple of 2π to each phase value.
4.3 Classic Fourier Series
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The classic Fourier series as derived originally expressed a periodic signal (period T ) in
terms of harmonically related sines and cosines.
The complex Fourier series and the sine-cosine series are identical, each
representing a signal's spectrum. The Fourier coefficients, ak and bk, express the
real and imaginary parts respectively of the spectrum while the coefficients c k of the
complex Fourier series express the spectrum as a magnitude and phase. Equating the
classic Fourier series (4.11) to the complex Fourier series (4.1), an extra factor of two
and complex conjugate become necessary to relate the Fourier coefficients in each.
Exercise 4.3.1
Derive this relationship between the coefficients of the two Fourier
series. Just as with the complex Fourier series, we can find the
Fourier coefficients using the orthogonally properties of sinusoids.
Note that the cosine and sine of harmonically related frequencies,
even the same frequency, are orthogonal.
These orthogonality relations follow from the following important trigonometric
identities.
139
These identities allow you to substitute a sum of sines and/or cosines for a product of
them. Each term in the sum can be integrated by noticing one of two important
properties of sinusoids.
• The integral of a sinusoid over an integer number of periods equals zero.
• The integral of the square of a unit-amplitude sinusoid over a period T equals
.
To use these, let's, for example, multiply the Fourier series for a signal by the cosine of
the l
th
harmonic
and integrate. The idea is that, because integration is linear, the integration will sift out
all but the term involving al.
The first and third terms are zero; in the second, the only non-zero term in the sum
results when the indices k and l are equal (but not zero), in which case we obtain
. If k =0= l, we obtain a0T . Consequently,
All of the Fourier coefficients can be found similarly.
140
Exercise 4.3.2
The expression for a0 is referred to as the average value of s(t). Why?
Exercise 4.3.3
What is the Fourier series for a unit-amplitude square wave?
141
Example 4.2
Let's find the Fourier series representation for the half-wave
rectified sinusoid.
Begin with the sine terms in the series; to find bk we must calculate
the integral
Using our trigonometric identities turns our integral of a product of
sinusoids into a sum of integrals of individual sinusoids, which are
much easier to evaluate.
Thus,
On to the cosine terms. The average value, which corresponds to a0,
equals
. The remainder of the cosine coefficients are easy to fnd, but yield
the complicated result
Thus, the Fourier series for the half-wave rectified sinusoid has nonzero terms for the average, the fundamental, and the even
harmonics.
142
4.4 A Signal's Spectrum
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A periodic signal, such as the half-wave rectified sinusoid, consists of a sum of
elemental sinusoids. A plot of the Fourier coefficients as a function of the frequency
index, such as shown in Figure 4.3 (Fourier Series spectrum of a half-wave rectified
sine wave), displays the signal's spectrum. The word "spectrum" implies that the
independent variable, here k, corresponds somehow to frequency. Each coefficient is
directly related to a sinusoid having a frequency of
Thus, if we half-wave rectified a 1 kHz sinusoid, k =1 corresponds to 1 kHz, k =2 to 2
kHz, etc.
Figure 4.3 Fourier Series spectrum of a half-wave rectified sine wave The Fourier series spectrum of a
half-wave rectified sinusoid is shown. The index indicates the multiple of the fundamental frequency at
which the signal has energy.
A subtle, but very important, aspect of the Fourier spectrum is its uniqueness: You
can unambiguously find the spectrum from the signal (decomposition (4.15)) and the
signal from the spectrum (composition). Thus, any aspect of the signal can be found
from the spectrum and vice versa. A signal's frequency domain expression is its
spectrum. A periodic signal can be defined either in the time domain (as a function)
or in the frequency domain (as a spectrum).
A fundamental aspect of solving electrical engineering problems is whether the time
or frequency domain provides the most understanding of a signal's properties and the
simplest way of manipulating it. The uniqueness property says that either domain can
provide the right answer. As a simple example, suppose we want to know the
(periodic) signal's maximum value. Clearly the time domain provides the answer
directly. To use a frequency domain approach would require us to find the spectrum,
form the signal from the spectrum and calculate the maximum; we're back in the time
domain!
143
Another feature of a signal is its average power. A signal's instantaneous power is
defined to be its square. The average power is the average of the instantaneous
power over some time interval. For a periodic signal, the natural time interval is clearly
its period; for nonperiodic signals, a better choice would be entire time or time from
onset. For a periodic signal, the average power is the square of its root-mean-squared
(rms) value. We define the rms value of a periodic signal to be
(4.20)
and thus its average power is
Exercise 4.4.1
What is the rms value of the half-wave rectified sinusoid?
To find the average power in the frequency domain, we need to substitute the spectral
representation of the signal into this expression.
The square inside the integral will contain all possible pairwise products. However, the
orthogonality properties (4.12) say that most of these crossterms integrate to zero.
The survivors leave a rather simple expression for the power we seek.
Figure 4.4 Power Spectrum of a Half-Wave Rectified Sinusoid Power spectrum of a half-wave rectified
sinusoid.
144
It could well be that computing this sum is easier than integrating the signal's square.
Furthermore, the contribution of each term in the Fourier series toward representing
the signal can be measured by its contribution to the signal's average power. Thus, the
power contained in a signal at its kth harmonic is
. The power spectrum, Ps (k), such as shown inFigure 4.4, plots each harmonic's
contribution to the total power.
Exercise 4.4.2
In high-end audio, deviation of a sine wave from the ideal is
measured by the total harmonic distortion, which equals the total
power in the harmonics higher than the first compared to power in
the fundamental. Find an expression for the total harmonic
distortion for any periodic signal. Is this calculation most easily
performed in the time or frequency domain?
4.5 Fourier Series Approximation of Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
It is interesting to consider the sequence of signals that we obtain as we incorporate
more terms into the Fourier series approximation of the half-wave rectified sine wave
(Example 4.2). Define sK (t) to be the signal containing K +1 Fourier terms.
Figure 4.5shows how this sequence of signals portrays the signal more accurately as
more terms are added.
145
Figure 4.5 Fourier Series spectrum of a half-wave rectified sine wave The Fourier series spectrum of a
half-wave rectified sinusoid is shown in the upper portion. The index indicates the multiple of the
fundamental frequency at which the signal has energy. The cumulative effect of adding terms to the Fourier
series for the half-wave rectified sine wave is shown in the bottom portion. The dashed line is the actual
signal, with the solid line showing the finite series approximation to the indicated number of terms, K +1.
We need to assess quantitatively the accuracy of the Fourier series approximation so
that we can judge how rapidly the series approaches the signal. When we use a K +1term series, the error the difference between the signal and the K +1-term series
corresponds to the unused terms from the series.
To find the rms error, we must square this expression and integrate it over a period.
Again, the integral of most cross-terms is zero, leaving
146
Figure 4.6 shows how the error in the Fourier series for the half-wave rectified
sinusoid decreases as more terms are incorporated. In particular, the use of four
terms, as shown in the bottom plot of Figure 4.5, has a rms error (relative to the rms
value of the signal) of about 3%. The Fourier series in this case converges quickly to
the signal.
Figure 4.6 Approximation error for a half-wave rectified sinusoid : The rms error calculated according
to (4.25) is shown as a function of the number of terms in the series for the half-wave rectified sinusoid. The
error has been normalized by the rms value of the signal.
We can look at Figure 4.7 to see the power spectrum and the rms approximation error
for the square wave.
147
:
Figure 4.7 Power spectrum and approximation error for a square wave The upper plot shows the
power spectrum of the square wave, and the lower plot the rms error of the fnite-length Fourier series
approximation to the square wave. The asterisk denotes the rms error when the number of terms K in the
Fourier series equals 99.
Because the Fourier coefficients decay more slowly here than for the half-wave
rectified sinusoid, the rms error is not decreasing quickly. Said another way, the
square-wave's spectrum contains more power at higher frequencies than does the
half-wave-rectified sinusoid. This difference between the two Fourier series results
because the half-wave rectified sinusoid's Fourier coefficients are proportional to
while those of the square wave are proportional to
If fact, after 99 terms of the square wave's approximation, the error is bigger than 10
terms of the approximation for the half-wave rectified sinusoid. Mathematicians have
shown that no signal has an rms approximation error that decays more slowly than it
does for the square wave.
148
Exercise 4.5.1
Calculate the harmonic distortion for the square wave. More than just
decaying slowly, Fourier series approximation shown in Figure 4.8
(Fourier series approximation of a square wave) exhibits interesting
behavior.
Figure 4.8 Fourier series approximation of a square wave Fourier series approximation to
sq (t). The number of terms in the Fourier sum is indicated in each plot, and the square
wave is shown as a dashed line over two periods.
Although the square wave's Fourier series requires more terms for a given
representation accuracy, when comparing plots it is not clear that the two are equal.
Does the Fourier series really equal the square wave at all values of t? In particular, at
each step-change in the square wave, the Fourier series exhibits a peak followed by
rapid oscillations. As more terms are added to the series, the oscillations seem to
become more rapid and smaller, but the peaks are not decreasing. For the Fourier
series approximation for the half-wave rectified sinusoid (Figure 4.5: Fourier Series
spectrum of a half-wave rectified sine wave), no such behavior occurs. What is
happening?
149
Consider this mathematical question intuitively: Can a discontinuous function, like the
square wave, be expressed as a sum, even an infinite one, of continuous signals? One
should at least be suspicious, and in fact, it can't be thus expressed. This issue brought
Fourier9 much criticism from the French Academy of Science (Laplace, Lagrange,
Monge and LaCroix comprised the review committee) for several years after its
presentation on 1807. It was not resolved for almost a century, and its resolution is
interesting and important to understand from a practical viewpoint.
The extraneous peaks in the square wave's Fourier series never disappear; they are
termed Gibb'sphenomenon after the American physicist Josiah Willard Gibbs. They
occur whenever the signal is dis continuous, and will always be present whenever the
signal has jumps.
Let's return to the question of equality; how can the equal sign in the definition of the
Fourier series be justified? The partial answer is that pointwise each and every value
of t equality is not guaranteed. However, mathematicians later in the nineteenth
century showed that the rms error of the Fourier series was always zero.
What this means is that the error between a signal and its Fourier series
approximation may not be zero, but that its rms value will be zero! It is through the
eyes of the rms value that we redefine equality: The usual Definition of equality is
called pointwise equality: Two signals s1 (t), s2 (t) are said to be equal pointwise if s1
(t)= s2 (t) for all values of t. A new definition of equality is mean-square equality: Two
signals are said to be equal in the mean square if rms (s1 − s2)=0. For Fourier series,
Gibb's phenomenon peaks have finite height and zero width. The error differs from
zero only at isolated points whenever the periodic signal contains discontinuities and
equals about 9% of the size of the discontinuity. The value of a function at a finite set
of points does not affect its integral. This effect underlies the reason why defining the
value of a discontinuous function, like we refrained from doing in defining the step
function (Section 2.2.4: Unit Step), at its discontinuity is meaningless. Whatever you
pick for a value has no practical relevance for either the signal's spectrum or for how a
system responds to the signal. The Fourier series value "at" the discontinuity is the
average of the values on either side of the jump.
4.6 Encoding Information in the Frequency Domain
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
To emphasize the fact that every periodic signal has both a time and frequency
domain representation, we can exploit both to encode information into a signal. Refer
to the Fundamental Model of Communication Figure 1.4. We have an information
source, and want to construct a transmitter that produces a signal x (t). For the source,
let's assume we have information to encode every T seconds. For example, we want to
represent typed letters produced by an extremely good typist (a key is struck every T
seconds). Let's consider the complex Fourier series formula in the light of trying to
encode information.
150
We use a finite sum here merely for simplicity (fewer parameters to determine). An
important aspect of the spectrum is that each frequency component ck can be
manipulated separately: Instead of finding the Fourier spectrum from a time-domain
specification, let's construct it in the frequency domain by selecting the ck according to
some rule that relates coefficient values to the alphabet. In defining this rule, we want
to always create a real-valued signal x (t). Because of the Fourier spectrum's
properties (Property 4.1, p. 121), the spectrum must have conjugate symmetry. This
requirement means that we can only assign positive-indexed coefficients (positive
frequencies), with negative-indexed ones equaling the complex conjugate of the
corresponding positive-indexed ones.
Assume we have N letters to encode: {a1,...,aN }. One simple encoding rule could be to
make a single Fourier coefficient be non-zero and all others zero for each letter. For
example, if an occurs , we make cn = 1 and ck=0,
. In this way, the n
th
harmonic of the frequency
is used to represent a letter. Note that the bandwidth the range of frequencies
required for the encoding equals
Another possibility is to consider the binary representation of the letter's index. For
example, if the letter a13 occurs, converting 13 to its base 2 representation, we have
13 = 11012. We can use the pattern of zeros and ones to represent directly which
Fourier coefficients we "turn on" (set equal to one) and which we "turn off."
Exercise 4.6.1
Compare the bandwidth required for the direct encoding scheme (one
nonzero Fourier coefficient for each letter) to the binary number
scheme. Compare the bandwidths for a 128-letter alphabet. Since
both schemes represent information without loss we can determine
the typed letter uniquely from the signal's spectrum both are viable.
Which makes more efficient use of bandwidth and thus might be
preferred?
151
Exercise 4.6.2
Can you think of an information-encoding scheme that makes even
more efficient use of the spectrum? In particular, can we use only one
Fourier coefficient to represent N letters uniquely? We can create an
encoding scheme in the frequency domain to represent an alphabet
of letters. But, as this information-encoding scheme stands, we can
represent one letter for all time. However, we note that the Fourier
coefficients depend only on the signal's characteristics over a single
period. We could change the signal's spectrum every T as each letter
is typed. In this way, we turn spectral coefficients on and of as letters
are typed, thereby encoding the entire typed document. For the
receiver (see the Fundamental Model of Communication (Figure 1.4))
to retrieve the typed letter, it would simply use the Fourier formula
for the complex Fourier spectrum for each T -second interval to
determine what each typed letter was. Figure 4.9 (Encoding Signals)
shows such a signal in the time-domain.
Figure 4.9 Encoding Signals The encoding of signals via the Fourier spectrum is shown over three
"periods." In this example, only the third and fourth harmonics are used, as shown by the spectral
magnitudes corresponding to each T -second interval plotted below the waveforms. Can you determine the
phase of the harmonics from the waveform?
In this Fourier-series encoding scheme, we have used the fact that spectral coefficients
can be independently specified and that they can be uniquely recovered from the
time-domain signal over one "period." Do note that the signal representing the entire
document is no longer periodic. By understanding the Fourier series' properties (in
particular that coefficients are determined only over a T -second interval, we can
construct a communications system. This approach represents a simplification of how
modern modems represent text that they transmit over telephone lines.
152
4.7 Filtering Periodic Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The Fourier series representation of a periodic signal makes it easy to determine how
a linear, time-invariant filter reshapes such signals in general. The fundamental
property of a linear system is that its input-output relation obeys superposition: L
(a1s1 (t)+ a2s2 (t)) = a1L (s1 (t)) + a2L (s2 (t)). Because the Fourier series represents a
periodic signal as a linear combination of complex exponentials, we can exploit the
superposition property. Furthermore, we found for linear circuits that their output to
a complex exponential input is just the frequency response evaluated at the signal's
frequency times the complex exponential. Said mathematically, if
, then the output
because
. Thus, if x(t) is periodic thereby
having a Fourier series, a linear circuit's output to this signal will be the superposition
of the output to each component.
Thus, the output has a Fourier series, which means that it too is periodic. Its Fourier
coefficients equal ckH (
) . To obtain the spectrum of the output, we simply multiply the input spectrum
by the frequency response. The circuit modifies the magnitude and phase of each
Fourier coefficient. Note especially that while the Fourier coefficients do not depend
on the signal's period, the circuit's transfer function does depend on frequency, which
means that the circuit's output will differ as the period varies.
153
Figure 4.10 Filtering a periodic signal A periodic pulse signal, such as shown on the left part ( =0.2), serves
as the input to an RC lowpass filter. The input's period was 1 ms (millisecond). The filter’s cutoff frequency
was set to the various values indicated in the top row, which display the output signal's spectrum and the
filter’s transfer function. The bottom row shows the output signal derived from the Fourier series
coefficients shown in the top row. (a) Periodic pulse signal (b) Top plots show the pulse signal's spectrum for
various cutoff frequencies. Bottom plots show the filter’s output signals.
Example 4.3
The periodic pulse signal shown on the left above serves as the input
to a RC-circuit that has the transfer function (calculated elsewhere
(Figure 3.30: Magnitude and phase of the transfer function))
Figure 4.10 shows the output changes as we vary the filter’s cutoff frequency. Note
how the signal's spectrum extends well above its fundamental frequency. Having a
cutoff frequency ten times higher than the fundamental does perceptibly change the
output waveform, rounding the leading and trailing edges. As the cutoff frequency
decreases (center, then left), the rounding becomes more prominent, with the
leftmost waveform showing a small ripple.
154
Exercise 4.7.1
What is the average value of each output waveform? The correct
answer may surprise you.
This example also illustrates the impact a lowpass filter can have on a waveform. The
simple RC filter used here has a rather gradual frequency response, which means that
higher harmonics are smoothly suppressed. Later, we will describe filters that have
much more rapidly varying frequency responses, allowing a much more dramatic
selection of the input's Fourier coefficients.
More importantly, we have calculated the output of a circuit to a periodic input
without writing, much less solving, the differential equation governing the circuit's
behavior. Furthermore, we made these calculations entirely in the frequency domain.
Using Fourier series, we can calculate how any linear circuit will respond to a periodic
input.
4.8 Derivation of the Fourier Transform
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Fourier series clearly open the frequency domain as an interesting and useful way of
determining how circuits and systems respond to periodic input signals. Can we use
similar techniques for nonperiodic signals? What is the response of the filter to a
single pulse? Addressing these issues requires us to find the Fourier spectrum of all
signals, both periodic and nonperiodic ones. We need a Definition for the Fourier
spectrum of a signal, periodic or not. This spectrum is calculated by what is known as
the Fourier transform.
Let sT (t) be a periodic signal having period T. We want to consider what happens to
this signal's spectrum as we let the period become longer and longer. We denote the
spectrum for any assumed value of the period by ck (T). We calculate the spectrum
according to the familiar formula
where we have used a symmetric placement of the integration interval about the
origin for subsequent deriva tional convenience. Let f be a fxed frequency equaling Tk
; we vary the frequency index k proportionally as we increase the period. Define
making the corresponding Fourier series
155
As the period increases, the spectral lines become closer together, becoming a
continuum. Therefore,
with
S (f) is the Fourier transform of s (t) (the Fourier transform is symbolically denoted by
the uppercase version of the signal's symbol) and is defined for any signal for which
the integral ((4.33)) converges.
Example 4.4
Let's calculate the Fourier transform of the pulse signal (Section
2.2.5: Pulse), p(t).
Note how closely this result resembles the expression for Fourier series coefficients of
the Figure 4.10.
Figure 4.11 Spectrum The upper plot shows the magnitude of the Fourier series spectrum for the case of T
=1 with the Fourier transform of p(t) shown as a dashed line. For the bottom panel, we expanded the period
to T =5, keeping the pulse's duration fixed at 0.2, and computed its Fourier series coefficients.
156
Figure 4.11 (Spectrum) shows how increasing the period does indeed lead to a
continuum of coefficients, and that the Fourier transform does correspond to what
the continuum becomes. The quantity
has a special name, the sinc (pronounced "sink") function, and is denoted by sinc (t).
Thus, the magnitude of the pulse's Fourier transform equals |Δsinc (πfΔ) |.
The Fourier transform relates a signal's time and frequency domain representations
to each other. The direct Fourier transform (or simply the Fourier transform)
calculates a signal's frequency domain representation from its time-domain variant
((4.34)). The inverse Fourier transform finds the time-domain representation from the
frequency domain. Rather than explicitly writing the required integral, we often
symbolically express these transform calculations as F (s) and F
−1
We must have s (t)= F (F (s (t))) and S (f)= F(F
valid with minor exceptions.
−1
−1
(S), respectively.
(S (f)), and these results are indeed
Note: Recall that the Fourier series for a square wave gives a value for the
signal at the discontinuities equal to the average value of the jump. This value
may difer from how the signal is defined in the time domain, but being
unequal at a point is indeed minor.
Showing that you "get back to where you started" is difficult from an analytic
viewpoint, and we won't try here. Note that the direct and inverse transforms differ
only in the sign of the exponent.
Exercise 4.8.1
The differing exponent signs means that some curious results occur
when we use the wrong sign. What is F (S (f))? In other words, use the
wrong exponent sign in evaluating the inverse Fourier transform.
Properties of the Fourier transform and some useful transform pairs are provided in
the accompanying tables (Table 4.1 and Table 4.2). Especially important among these
properties is Parseval's Theorem, which states that power computed in either
domain equals the power in the other.
157
Of practical importance is the conjugate symmetry property: When s (t) is real-valued,
the spectrum at negative frequencies equals the complex conjugate of the spectrum
at the corresponding positive frequencies. Consequently, we need only plot the
positive frequency portion of the spectrum (we can easily determine the remainder of
the spectrum).
Exercise 4.8.2
How many Fourier transform operations need to be applied to get the
original signal back:
Note that the mathematical relationships between the time domain and frequency
domain versions of the same signal are termed transforms. We are transforming (in
the nontechnical meaning of the word) a signal from one representation to another.
We express Fourier transform pairs as (s (t) ↔ S (f)). A signal's time and frequency
domain representations are uniquely related to each other. A signal thus "exists" in
both the time and frequency domains, with the Fourier transform bridging between
the two. We can define an information carrying signal in either the time or frequency
domains; it behooves the wise engineer to use the simpler of the two.
A common misunderstanding is that while a signal exists in both the time and
frequency domains, a single formula expressing a signal must contain only time or
frequency: Both cannot be present simultaneously. This situation mirrors what
happens with complex amplitudes in circuits: As we reveal how communications
systems work and are designed, we will define signals entirely in the frequency
domain without explicitly finding their time domain variants. This idea is shown in
another module (Section 4.6) where we define Fourier series coefficients according to
letter to be transmitted. Thus, a signal, though most familiarly defined in the timedomain, really can be defined equally as well (and sometimes more easily) in the
frequency domain. For example, impedances depend on frequency and the time
variable cannot appear.
We will learn (Section 4.9) that finding a linear, time-invariant system's output in the
time domain can be most easily calculated by determining the input signal's spectrum,
performing a simple calculation in the frequency domain, and inverse transforming
the result. Furthermore, understanding communications and information processing
systems requires a thorough understanding of signal structure and of how systems
work in both the time and frequency domains.
The only difficulty in calculating the Fourier transform of any signal occurs when we
have periodic signals (in either domain). Realizing that the Fourier series is a special
case of the Fourier transform, we simply calculate the Fourier series coefficients
instead, and plot them along with the spectra of nonperiodic signals on the same
frequency axis.
158
Table 4.1 Short Table of Fourier Transform Pairs
159
Time-Domain
Frequency Domain
Linearity
Conjugate
Symmetry
Even Symmetry
Odd Symmetry
Scale Change
Time Delay
Complex
Modulation
Amplitude
Modulation by
Cosine
Amplitude
Modulation by
Sine
Differentiation
Multiplication
by t
Area
Value at Origin
Parseval"s
Theorem
Table 4.2 Fourier Transform Properties
Example 4.5
In communications, a very important operation on a signal s (t) is to
amplitude modulate it. Using this operation more as an example
rather than elaborating the communications aspects here, we want to
compute the Fourier transform the spectrum of
Thus,
160
For the spectrum of cos (2πfct), we use the Fourier series. Its period is
and its only nonzero Fourier coefficients are
The second term is not periodic unless s (t) has the same period as the sinusoid.
Using Euler's relation, the spectrum of the second term can be derived as
Using Euler's relation for the cosine,
Exploiting the uniqueness property of the Fourier transform, we have
This component of the spectrum consists of the original signal's spectrum delayed and
advanced in frequency. The spectrum of the amplitude modulated signal is shown in
Figure 4.12.
Figure 4.12 Spectrum of the amplitude modulated signal A signal which has a triangular shaped
spectrum is shown in the top plot. Its highest frequency the largest frequency containing power is W Hz.
Once amplitude modulated, the resulting spectrum has "lines" corresponding to the Fourier series
components at ± (fc) and the original triangular spectrum shifted to components at ± (fc) and scaled by .
Note how in this figure the signal s (t) is defined in the frequency domain. To find its
time domain representation, we simply use the inverse Fourier transform.
161
Exercise 4.8.3
What is the signal s (t) that corresponds to the spectrum shown in the
upper panel of Figure 4.12?
Exercise 4.8.4
What is the power in x (t), the amplitude-modulated signal? Try the
calculation in both the time and frequency domains.
In this example, we call the signal s (t) a baseband signal because its power is
contained at low frequencies. Signals such as speech and the Dow Jones averages are
baseband signals. The baseband signal's bandwidth equals W , the highest frequency
at which it has power. Since x (t)'s spectrum is confned to a frequency band not close
to the origin (we assume fc» W ) , we have a bandpass signal. The bandwidth of a
bandpasssignal is not its highest frequency, but the range of positive frequencies
where the signal has power. Thus, in this example, the bandwidth is 2W Hz. Why a
signal's bandwidth should depend on its spectral shape will become clear once we
develop communications systems.
4.9 Linear Time Invariant Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we apply a periodic input to a linear, time-invariant system, the output is
periodic and has Fourier series coefficients equal to the product of the system's
frequency response and the input's Fourier coefficients (Filtering Periodic Signals
(4.27)). The way we derived the spectrum of non-periodic signal from periodic ones
makes it clear that the same kind of result works when the input is not periodic: If x (t)
serves as the input to a linear, time-invariant system having frequency response
H (f), the spectrum of the output isX (f) H (f).
162
Example 4.6
Let's use this frequency-domain input-output relationship for
linear, time-invariant systems to find a formula for the RC-circuit's
response to a pulse input. We have expressions for the input's
spectrum and the system's frequency response.
Thus, the output's Fourier transform equals
You won't find this Fourier transform in our table, and the required
integral is difficult to evaluate as the expression stands. This
situation requires cleverness and an understanding of the Fourier
transform's properties. In particular, recall Euler's relation for the
sinusoidal term and note the fact that multiplication by a complex
exponential in the frequency domain amounts to a time delay. Let's
momentarily make the expression for Y (f) more complicated.
Consequently,
The table of Fourier transform properties (Figure 4.13) suggests thinking about this
expression as a product of terms.
• Multiplication by
means integration.
• Multiplication by the complex exponential e
the time domain.
−(j2πfΔ)
means delay by Δ seconds in
163
−(j2πfΔ)
• The term 1 − e
from its original.
means, in the time domain, subtract the time-delayed signal
• The inverse transform of the frequency response is
We can translate each of these frequency-domain products into time-domain
operations in any order we like because the order in which multiplications occur
doesn't affect the result. Let's start with the product of
(integration in the time domain) and the transfer function:
The middle term in the expression for Y (f) consists of the difference of two terms: the
constant 1 and the complex exponential e. Because of the Fourier transform's
linearity, we simply subtract the results.
Note that in delaying the signal how we carefully included the unit step. The second
term in this result does not begin until t =Δ. Thus, the waveforms shown in the
Filtering Periodic Signals (Figure 4.10: Filtering a periodic signal) example mentioned
above are exponentials. We say that the time constant of an exponentially decaying
signal equals the time it takes to decrease by
of its original value. Thus, the time-constant of the rising and falling portions of the
output equal the product of the circuit's resistance and capacitance.
Exercise 4.9.1
Derive the filter’s output by considering the terms in (4.41) in the
order given. Integrate last rather than first. You should get the same
answer.
In this example, we used the table extensively to find the inverse Fourier transform,
relying mostly on what multiplication by certain factors, like
and
, meant. We essentially treated multiplication by these factors as if they were transfer
functions of some fictitious circuit. The transfer function
corresponded to a circuit that integrated, and
to one that delayed. We even implicitly interpreted.
164
the circuit's transfer function as the input's spectrum! This approach to finding inverse
transforms breaking down a complicated expression into products and sums of
simple components is the engineer's way of breaking down the problem into several
subproblems that are much easier to solve and then gluing the results together. Along
the way we may make the system serve as the input, but in the rule Y (f)= X (f) H (f),
which term is the input and which is the transfer function is merely a notational
matter (we labeled one factor with an X and the other with an H).
4.9.1 Transfer Functions
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The notion of a transfer function applies well beyond linear circuits. Although we don't
have all we need to demonstrate the result as yet, all linear, time-invariant systems
have a frequency-domain input-output relation given by the product of the input's
Fourier transform and the system's transfer function. Thus, linear circuits are a special
case of linear, time-invariant systems. As we tackle more sophisticated problems in
transmitting, manipulating, and receiving information, we will assume linear systems
having certain properties (transfer functions) without worrying about what circuit has
the desired property. At this point, you may be concerned that this approach is glib,
and rightly so. Later we'll show that by involving software that we really don't need to
be concerned about constructing a transfer function from circuit elements and opamps.
4.9.2 Commutative Transfer Functions
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Another interesting notion arises from the commutative property of multiplication
(exploited in an example above (Example 4.6)): We can rather arbitrarily choose an
order in which to apply each product. Consider a cascade of two linear, time-invariant
systems. Because the Fourier transform of the first system's output is X (f) H1 (f) and it
serves as the second system's input, the cascade's output spectrum is X (f) H1 (f) H2
(f). Because this product also equals X (f) H2 (f) H1 (f), the cascade having the linear
systems in the opposite order yields the same result. Furthermore, the cascade
acts like a single linear system, having transfer function H1 (f) H2 (f). This result applies
to other configurations of linear, time-invariant systems as well; see this Frequency
Domain Problem (Problem 4.13). Engineers exploit this property by determining what
transfer function they want, then breaking it down into components arranged
according to standard confgurations. Using the fact that op-amp circuits can be
connected in cascade with the transfer function equaling the product of its
component's transfer function (see this analog signal processing problem (Problem
3.44)), we find a ready way of realizing designs. We now understand why op-amp
implementations of transfer functions are so important.
165
4.10 Modeling the Speech Signal
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 4.13 Vocal Tract The vocal tract is shown in cross-section. Air pressure produced by the lungs forces
air through the vocal cords that, when under tension, produce puffs of air that excite resonances in the
vocal and nasal cavities. What are not shown are the brain and the musculature that control the entire
speech production process.
166
Figure 4.14 Model of the Vocal Tract The systems model for the vocal tract. The signals l (t), pT (t), and s (t)
are the air pressure provided by the lungs, the periodic pulse output provided by the vocal cords, and the
speech output respectively. Control signals from the brain are shown as entering the systems from the top.
Clearly, these come from the same source, but for modeling purposes we describe them separately since
they control different aspects of the speech signal.
The information contained in the spoken word is conveyed by the speech signal.
Because we shall analyze several speech transmission and processing schemes, we
need to understand the speech signal's structure what's special about the speech
signal and how we can describe and model speech production. This modeling effort
consists of finding a system's description of how relatively unstructured signals,
arising from simple sources, are given structure by passing them through an
interconnection of systems to yield speech. For speech and for many other situations,
system choice is governed by the physics underlying the actual production process.
Because the fundamental equation of acoustics the wave equation applies here and is
linear, we can use linear systems in our model with a fair amount of accuracy. The
naturalness of linear system models for speech does not extend to other situations. In
many cases, the underlying mathematics governed by the physics, biology, and/or
chemistry of the problem are nonlinear, leaving linear systems models as
approximations. Nonlinear models are far more difcult at the current state of
knowledge to understand, and information engineers frequently prefer linear models
because they provide a greater level of comfort, but not necessarily a sufcient level of
accuracy.
Figure 4.13 shows the actual speech production system and Figure 4.14 shows the
model speech production system. The characteristics of the model depends on
whether you are saying a vowel or a consonant. We concentrate first on the vowel
production mechanism. When the vocal cords are placed under tension by the
surrounding musculature, air pressure from the lungs causes the vocal cords to
vibrate. To visualize this efect, take a rubber band and hold it in front of your lips. If
held open when you blow through it, the air passes through more or less freely; this
situation corresponds to "breathing mode". If held tautly and close together, blowing
through the opening causes the sides of the rubber band to vibrate. This effect works
best with a wide rubber band. You can imagine what the airflow is like on the opposite
side of the rubber band or the vocal cords. Your lung power is the simple source
referred to earlier; it can be modeled as a constant supply of air pressure. The vocal
cords respond to this input by vibrating, which means the output of this system is
some periodic function.
167
Exercise 4.10.1
Note that the vocal cord system takes a constant input and produces
a periodic airflow that corresponds to its output signal. Is this system
linear or nonlinear? Justify your answer.
Singers modify vocal cord tension to change the pitch to produce the desired musical
note. Vocal cord tension is governed by a control input to the musculature; in system's
models we represent control inputs as signals coming into the top or bottom of the
system. Certainly in the case of speech and in many other cases as well, it is the
control input that carries information, impressing it on the system's output. The
change of signal structure resulting from varying the control input enables
information to be conveyed by the signal, a process generically known as modulation.
In singing, musicality is largely conveyed by pitch; in western speech, pitch is much
less important. A sentence can be read in a monotone fashion without completely
destroying the information expressed by the sentence. However, the difference
between a statement and a question is frequently expressed by pitch changes. For
example, note the sound differences between "Let's go to the park." and "Let's go to
the park?";
For some consonants, the vocal cords vibrate just as in vowels. For example, the socalled nasal sounds "n" and "m" have this property. For others, the vocal cords do not
produce a periodic output. Going back to mechanism, when consonants such as "f"
are produced, the vocal cords are placed under much less tension, which results in
turbulent flow. The resulting output airflow is quite erratic, so much so that we
describe it as being noise. We define noise carefully later when we delve into
communication problems.
The vocal cords' periodic output can be well described by the periodic pulse train pT (t)
as shown in the periodic pulse signal (Figure 4.1), with T denoting the pitch period. The
spectrum of this signal (4.9) contains harmonics of the frequency
, what is known as the pitch frequency or the fundamental frequency F0. The
primary difference between adult male and female/prepubescent speech is pitch.
Before puberty, pitch frequency for normal speech ranges between 150-400 Hz for
both males and females. After puberty, the vocal cords of males undergo a physical
change, which has the effect of lowering their pitch frequency to the range 80-160 Hz.
If we could examine the vocal cord output, we could probably discern whether the
speaker was male or female. This diference is also readily apparent in the speech
signal itself.
To simplify our speech modeling efort, we shall assume that the pitch period is
constant. With this simplifcation, we collapse the vocal-cord-lung system as a simple
source that produces the periodic pulse signal (Figure 4.14). The sound pressure signal
thus produced enters the mouth behind the tongue, creates acoustic disturbances,
and exits primarily through the lips and to some extent through the nose. Speech
specialists tend to name the mouth, tongue, teeth, lips, and nasal cavity the vocal
tract. The physics governing the sound disturbances produced in the vocal tract and
168
those of an organ pipe are quite similar. Whereas the organ pipe has the simple
physical structure of a straight tube, the cross-section of the vocal tract "tube" varies
along its length because of the positions of the tongue, teeth, and lips. It is these
positions that are controlled by the brain to produce the vowel sounds. Spreading the
lips, bringing the teeth together, and bringing the tongue toward the front portion of
the roof of the mouth produces the sound "ee." Rounding the lips, spreading the
teeth, and positioning the tongue toward the back of the oral cavity produces the
sound "oh." These variations result in a linear, time-invariant system that has a
frequency response typifed by several peaks, as shown in Figure 4.15.
Figure 4.15 Speech Spectrum The ideal frequency response of the vocal tract as it produces the sounds
"oh" and "ee" are shown on the top left and top right, respectively. The spectral peaks are known as
formants, and are numbered consecutively from low to high frequency. The bottom plots show speech
waveforms corresponding to these sounds.
These peaks are known as formants. Thus, speech signal processors would say that
the sound "oh" has a higher first formant frequency than the sound "ee," with F2
being much higher during "ee." F2 and F3 (the second and third formants) have more
energy in "ee" than in "oh." Rather than serving as a filter, rejecting high or low
frequencies, the vocal tract serves to shape the spectrum of the vocal cords. In the
time domain, we have a periodic signal, the pitch, serving as the input to a linear
system. We know that the output the speech signal we utter and that is heard by
others and ourselves will also be periodic. Example time-domain speech signals are
shown in Figure 4.15, where the periodicity is quite apparent.
169
Exercise 4.10.2
From the waveform plots shown in Figure 4.15, determine the pitch
period and the pitch frequency. Since speech signals are periodic,
speech has a Fourier series representation given by a linear circuit's
response to a periodic signal (4.27). Because the acoustics of the
vocal tract are linear, we know that the spectrum of the output equals
the product of the pitch signal's spectrum and the vocal tract's
frequency response. We thus obtain the fundamental model of
speech production.
Here, HV (f) is the transfer function of the vocal tract system. The Fourier series for the
vocal cords' output, derived in this equation (p. 122), is
and is plotted on the top in Figure 4.16. If we had, for example, a male speaker with
about a 110 Hz pitch (
) saying the vowel "oh", the spectrum of his speech predicted by our model is shown
in Figure 4.16.
170
Figure 4.16 voice spectrum The vocal tract's transfer function, shown as the thin, smooth line, is
superimposed on the spectrum of actual male speech corresponding to the sound "oh." The pitch lines
corresponding to harmonics of the pitch frequency are indicated. (a) The vocal cords' output spectrum P T (f).
(b) The vocal tract's transfer function, HV (f) and the speech spectrum.
The model spectrum idealizes the measured spectrum, and captures all the important
features. The measured spectrum certainly demonstrates what are known as pitch
lines, and we realize from our model that they are due to the vocal cord's periodic
excitation of the vocal tract. The vocal tract's shaping of the line spectrum is clearly
evident, but difficult to discern exactly, especially at the higher frequencies. The model
transfer function for the vocal tract makes the formants much more readily evident.
Exercise 4.10.3
The Fourier series coefficients for speech are related to the vocal
tract's transfer function only at the frequencies
see previous result (4.9). Would male or female speech tend to
T have a more clearly identifiable formant structure when its spectrum is computed?
Consider, for example, how the spectrum shown on the right in Figure 4.16 would
change if the pitch were twice as high (≈ (300) Hz).
171
When we speak, pitch and the vocal tract's transfer function are not static; they
change according to their control signals to produce speech. Engineers typically
display how the speech spectrum changes over time with what is known as a
Spectrograms (Page 224) Figure 4.17. Note how the line spectrum, which indicates
how the pitch changes, is visible during the vowels, but not during the consonants (like
the ce in "Rice").
:
Figure 4.17 spectrogram Displayed is the spectrogram of the author saying "Rice University." Blue
indicates low energy portion of the spectrum, with red indicating the most energetic portions. Below the
spectrogram is the time-domain speech signal, where the periodicities can be seen.
The fundamental model for speech indicates how engineers use the physics
underlying the signal generation process and exploit its structure to produce a
systems model that suppresses the physics while emphasizing how the signal is
"constructed." From everyday life, we know that speech contains a wealth of
information. We want to determine how to transmit and receive it. Efficient and
effective speech transmission requires us to know the signal's properties and its
structure (as expressed by the fundamental model of speech production). We see
from Figure 4.17, for example, that speech contains significant energy from zero
frequency up to around 5 kHz.
Effective speech transmission systems must be able to cope with signals having this
bandwidth. It is interesting that one system that does not support this 5 kHz
bandwidth is the telephone: Telephone systems act like a bandpass filter passing
energy between about 200 Hz and 3.2 kHz. The most important consequence of this
filtering is the removal of high frequency energy. In our sample utterance, the "ce"
172
sound in "Rice"" contains most of its energy above 3.2 kHz; this filtering effect is why it
is extremely difficult to distinguish the sounds "s" and "f" over the telephone. Try this
yourself: Call a friend and determine if they can distinguish between the words "six"
and "fx". If you say these words in isolation so that no context provides a hint about
which word you are saying, your friend will not be able to tell them apart. Radio does
support this bandwidth (see more about Signal-to-Noise Ratio of an AmplitudeModulated Signal (Page 278) ).
Efficient speech transmission systems exploit the speech signal's special structure:
What makes speech speech? You can conjure many signals that span the same
frequencies as speech car engine sounds, violin music, dog barks but don't sound at
all like speech. We shall learn later that transmission of any 5 kHz bandwidth signal
requires about 80 kbps (thousands of bits per second) to transmit digitally. Speech
signals can be transmitted using less than 1 kbps because of its special structure. To
reduce the "digital bandwidth" so drastically means that engineers spent many years
to develop signal processing and coding methods that could capture the special
characteristics of speech without destroying how it sounds. If you used a speech
transmission system to send a violin sound, it would arrive horribly distorted; speech
transmitted the same way would sound fine.
Exploiting the special structure of speech requires going beyond the capabilities of
analog signal processing systems. Many speech transmission systems work by finding
the speaker's pitch and the formant frequencies. Fundamentally, we need to do more
than filtering to determine the speech signal's structure; we need to manipulate
signals in more ways than are possible with analog systems. Such flexibility is
achievable (but not without some loss) with programmable digital systems.
173
4.11 Frequency Domain Problems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Problem 4.1: Simple Fourier Series
Find the complex Fourier series representations of the following
signals without explicitly calculating Fourier integrals. What is the
signal's period in each case?
1.
2.
3.
4.
s (t) = sin(t)
s (t) = sin2 (t)
s (t) = cos(t) + 2cos (2t)
s (t) = cos (2t) cos (t)
5.
6. s (t) given by the depicted waveform (Figure 4.18)
Figure 4.18 depicted waveform
174
Problem 4.2: Fourier Series
Find the Fourier series representation for the following periodic
signals (Figure 4.19). For the third signal, find the complex Fourier
series for the triangle wave without performing the usual Fourier
integrals. Hint:
How is this signal related to one for which you already have the
series?
Figure 4.19 Fourier Series
175
Problem 4.3: Phase Distortion
We can learn about phase distortion by returning to circuits and
investigate the following circuit (Figure 4.20).
Figure 4.20 Phase Distortion
1. Find this filter’s transfer function.
2. Find the magnitude and phase of this transfer function. How
would you characterize this circuit?
3. Let vin (t) be a square-wave of period T. What is the Fourier series
for the output voltage?
4. Use Matlab to find the output's waveform for the cases T =0.01
and T =2. What value of T delineates the two kinds of results you
found? The software in fourier2.m might be useful.
5. Instead of the depicted circuit, the square wave is passed through a
system that delays its input, which applies a linear phase shift to
the signal's spectrum. Let the delay τ be
. Use the transfer function of a delay to compute using Matlab the
Fourier series of the output. Show that the square wave is indeed
delayed.
176
Problem 4.4: Approximating Periodic
Signals
Often, we want to approximate a reference signal by a somewhat
simpler signal. To assess the quality of an approximation, the most
frequently used error measure is the mean-squared error. For a
periodic signal s (t),
where s (t) is the reference signal and
its approximation. One convenient way of finding approximations for
periodic signals is to truncate their Fourier series.
The point of this problem is to analyze whether this approach is the
best (i.e., always minimizes the mean-squared error).
1. Find a frequency-domain expression for the approximation error
when we use the truncated Fourier series as the approximation.
2. Instead of truncating the series, let's generalize the nature of the
approximation to including any set of 2K +1 terms: We'll always
include the c0 and the negative indexed term corresponding to ck.
What selection of terms minimizes the mean-squared error? Find
an expression for the mean-squared error resulting from your
choice.
3. Find the Fourier series for the depicted signal (Figure 4.21). Use
Matlab to find the truncated approximation and best
approximation involving two terms. Plot the mean-squared error
as a function of K for both approximations.
Figure 4.21 Approximating Periodic Signals
177
Problem 4.5: Long, Hot Days
The daily temperature is a consequence of several effects, one of
them being the sun's heating. If this were the dominant effect, then
daily temperatures would be proportional to the number of daylight
hours. The plot (Figure 4.22) shows that the average daily high
temperature does not behave that way.
Figure 4.22 Long, Hot Days
In this problem, we want to understand the temperature component
of our environment using Fourier series and linear system theory.
The file temperature.mat contains these data (daylight hours in the
first row, corresponding average daily highs in the second) for
Houston, Texas.
1. Let the length of day serve as the sole input to a system having an
output equal to the average daily temperature. Examining the
plots of input and output, would you say that the system is linear
or not? How did you reach you conclusion?
2. Find the first five terms (c0, ... , c4) of the complex Fourier series
for each signal. Use the following formula that approximates the
integral required to find the Fourier coefficients.
3. What is the harmonic distortion in the two signals? Exclude c0
from this calculation.
4. Because the harmonic distortion is small, let's concentrate only
on the first harmonic. What is the phase shift between input and
output signals?
5. Find the transfer function of the simplest possible linear model
that would describe the data. Characterize and interpret the
178
structure of this model. In particular, give a physical explanation
for the phase shift.
6. Predict what the output would be if the model had no phase shift.
Would days be hotter? If so, by how much?
Problem 4.6: Fourier Transform Pairs
Find the Fourier or inverse Fourier transform of the following.
Problem 4.7: Duality in Fourier
Transforms
"Duality" means that the Fourier transform and the inverse Fourier
transform are very similar. Consequently, the waveform s (t) in the
time domain and the spectrum s (f) have a Fourier transform and an
inverse Fourier transform, respectively, that are very similar.
1. Calculate the Fourier transform of the signal shown below (Figure
4.23(a)).
2. Calculate the inverse Fourier transform of the spectrum shown
below (Figure 4.23(b)).
3. How are these answers related? What is the general relationship
between the Fourier transform of s (t) and the inverse transform
of s (f)?
Figure 4.23 Duality in Fourier Transforms
179
Problem 4.8: Spectra of Pulse Sequences
Pulse sequences occur often in digital communication and in other
fields as well. What are their spectral properties?
1. Calculate the Fourier transform of the single pulse shown below
(Figure 4.24(a)).
2. Calculate the Fourier transform of the two-pulse sequence shown
below (Figure 4.24(b)).
3. Calculate the Fourier transform for the ten-pulse sequence shown
in below (Figure 4.24(c)). You should look for a general expression
that holds for sequences of any length.
4. Using Matlab, plot the magnitudes of the three spectra. Describe
how the spectra change as the number of repeated pulses
increases.
Figure 4.24 Spectra of Digital Communication Signals 1
One way to represent bits with signals is shown in Figure 4.25. If the
value of a bit is a "1", it is represented by a positive pulse of duration
T. If it is a "0", it is represented by a negative pulse of the same
duration. To represent a sequence of bits, the appropriately chosen
pulses are placed one after the other.
180
Figure 4.25 Spectra of Digital Communication Signals 2
1. What is the spectrum of the waveform that represents the
alternating bit sequence "...01010101..."?
2. This signal's bandwidth is defined to be the frequency range over
which 90% of the power is contained. What is this signal's
bandwidth?
3. Suppose the bit sequence becomes "...00110011..." Now what is
the bandwidth?
Problem 4.10: Lowpass Filtering a
Square Wave
Let a square wave (period T ) serve as the input to a first-order
lowpass system constructed as a RC filter. We want to derive an
expression for the time-domain response of the filter to this input.
1. First, consider the response of the filter to a simple pulse, having
unit amplitude and width
. Derive an expression for the filter's output to this pulse.
2. Noting that the square wave is a superposition of a sequence of
these pulses, what is the filter's response to the square wave?
3. The nature of this response should change as the relation between
the square wave's period and the filter's cutoff frequency change.
How long must the period be so that the response does not
achieve a relatively constant value between transitions in the
square wave? What is the relation of the filter's cutoff frequency to
the square wave's spectrum in this case?
Problem 4.11: Mathematics with Circuits
Simple circuits can implement simple mathematical operations, such
as integration and differentiation. We want to develop an active
circuit (it contains an op-amp) having an output that is proportional
to the integral of its input. For example, you could use an integrator
in a car to determine distance traveled from the speedometer.
1. What is the transfer function of an integrator?
2. Find an op-amp circuit so that its voltage output is proportional to
the integral of its input for all signals.
181
Problem 4.12: Where is that sound
coming from?
We determine where sound is coming from because we have two ears
and a brain. Sound travels at a relatively slow speed and our brain
uses the fact that sound will arrive at one ear before the other. As
shown here (Figure 4.26), a sound coming from the right arrives at
the left ear τ seconds after it arrives at the right ear.
Figure 4.26 Where is that sound coming from?
Once the brain fnds this propagation delay, it can determine the
sound direction. In an attempt to model what the brain might do, RU
signal processors want to design an optimal system that delays each
ear's signal by some amount then adds them together. Δl and Δr are
the delays applied to the left and right signals respectively. The idea
is to determine the delay values according to some criterion that is
based on what is measured by the two ears.
1. What is the transfer function between the sound signal s (t) and
the processor output y (t)?
2. One way of determining the delay τ is to choose Δl and Δr to
maximize the power in y (t). How are these maximum-power
processing delays related to τ?
182
Problem 4.13: Arrangements of Systems
Architecting a system of modular components means arranging them
in various confgurations to achieve some overall input-output
relation. For each of the following (Figure 4.27), determine the
overall transfer function between x (t) and y (t).
Figure 4.27 Arrangements of Systems
The overall transfer function for the cascade (first depicted system) is
particularly interesting. What does it say about the effect of the
ordering of linear, time-invariant systems in a cascade?
Problem 4.14: Filtering
Let the signal s (t)= sin(πt πt) be the input to a linear, time-invariant
filter having the transfer function shown below (Figure 4.28). Find
the expression for y (t), the filter's output.
Figure 4.28 Filtering
183
Problem 4.15: Circuits Filter!
A unit-amplitude pulse with duration of one second serves as the
input to an RC-circuit having transfer function
1. How would you categorize this transfer function: lowpass,
highpass, bandpass, other?
2. Find a circuit that corresponds to this transfer function.
3. Find an expression for the filter's output.
Problem 4.16: Reverberation
Reverberation corresponds to adding to a signal its delayed version.
1. Assuming τ represents the delay, what is the input-output relation
for a reverberation system? Is the system linear and timeinvariant? If so, find the transfer function; if not, what linearity or
time invariance criterion does reverberation violate.
2. A music group known as the ROwls is having trouble selling its
recordings. The record company's engineer gets the idea of
applying different delay to the low and high frequencies and
adding the result to create a new musical effect. Thus, the ROwls'
audio would be separated into two parts (one less than the
frequency f0, the other greater than f0), these would be delayed by
τl and τh respectively, and the resulting signals added. Draw a
block diagram for this new audio processing system, showing its
various components.
3. How does the magnitude of the system's transfer function depend
on the two delays?
Problem 4.17: Echoes in Telephone
Systems
A frequently encountered problem in telephones is echo. Here,
because of acoustic coupling between the ear piece and microphone
in the handset, what you hear is also sent to the person talking. That
person thus not only hears you, but also hears her own speech
delayed (because of propagation delay over the telephone network)
and attenuated (the acoustic coupling gain is less than one).
Furthermore, the same problem applies to you as well: The acoustic
coupling occurs in her handset as well as yours.
1. Develop a block diagram that describes this situation.
2. Find the transfer function between your voice and what the
listener hears.
3. Each telephone contains a system for reducing echoes using
electrical means. What simple system could null the echoes?
184
Problem 4.18: Effective Drug Delivery
In most patients, it takes time for the concentration of an
administered drug to achieve a constant level in the blood stream.
Typically, if the drug concentration in the patient's intravenous line
is Cdu (t), the concentration in the patient's blood stream is Cp( 1 −
e−(at))u (t).
1. Assuming the relationship between drug concentration in the
patient's drug and the delivered concentration can be described as
a linear, time-invariant system, what is the transfer function?
2. Sometimes, the drug delivery system goes awry and delivers drugs
with little control. What would the patient's drug concentration be
if the delivered concentration were a ramp? More precisely, if it
were Cdtu (t)?
3. A clever doctor wants to have the flexibility to slow down or speed
up the patient's drug concentration. In other words, the
concentration is to be Cp( 1 − e−(bt))u (t), with b bigger or smaller
than a. How should the delivered drug concentration signal be
changed to achieve this concentration profile?
Problem 4.19: Catching Speeders with
Radar
RU Electronics has been contracted to design a Doppler radar system.
Radar transmitters emit a signal that bounces of any conducting
object. Signal diferences between what is sent and the radar return is
processed and features of interest extracted. In Doppler systems, the
object's speed along the direction of the radar beam is the feature the
design must extract. The transmitted signal is a sinsusoid: x (t)= Acos
(2πfct). The measured return signal equals Bcos (2π ((fc + Δf) t + Ï•)),
where the Doppler offset frequency Δf equals 10v, where v is the car's
velocity coming toward the transmitter.
• Design a system that uses the transmitted and return signals as
inputs and produces Δf.
• One problem with designs based on overly simplistic design goals
is that they are sensitive to unmodeled assumptions. How would
you change your design, if at all, so that whether the car is going
away or toward the transmitter could be determined?
• Suppose two objects traveling different speeds provide returns.
How would you change your design, if at all, to accomodate
multiple returns?
185
Problem 4.20: Demodulating an AM
Signal
Let m (t) denote the signal that has been amplitude modulated.
Radio stations try to restrict the amplitude of the signal m (t) so that
it is less than one in magnitude. The frequency fcis very large
compared to the frequency content of the signal. What we are
concerned about here is not transmission, but reception.
1. The so-called coherent demodulator simply multiplies the signal x
(t) by a sinusoid having the same frequency as the carrier and
lowpass filters the result. Analyze this receiver and show that it
works. Assume the lowpass filter is ideal.
2. One issue in coherent reception is the phase of the sinusoid used
by the receiver relative to that used by the transmitter. Assuming
that the sinusoid of the receiver has a phase φ, how does the
output depend on φ? What is the worst possible value for this
phase?
3. The incoherent receiver is more commonly used because of the
phase sensitivity problem inherent in coherent reception. Here,
the receiver full-wave rectifes the received signal and lowpass
filters the result (again ideally). Analyze this receiver. Does its
output difer from that of the coherent receiver in a signifcant way?
186
Problem 4.21: Unusual Amplitude
Modulation
We want to send a band-limited signal having the depicted spectrum
(Figure 4.29(a)) with amplitude modulation in the usual way. I.B.
Different suggests using the square-wave carrier shown below
(Figure 4.29(b)). Well, it is different, but his friends wonder if any
technique can demodulate it.
1. Find an expression for X (f), the Fourier transform of the
modulated signal.
2. Sketch the magnitude of X (f), being careful to label important
magnitudes and frequencies.
3. What demodulation technique obviously works?
4. I.B. challenges three of his friends to demodulate x (t) some other
way. One friend suggests modulating x (t) with
, another wants to try modulating with
and the third thinks cos 2 will work. Sketch the magnitude of the
Fourier transform of the signal each student's approach produces.
Which student comes closest to recovering the original signal?
Why?
Figure 4.29 Unusual Amplitude Modulation
187
Problem 4.22: Sammy Falls Asleep...
While sitting in ELEC 241 class, he falls asleep during a critical time
when an AM receiver is being described. The received signal has the
form r (t)= A(1 + m (t)) cos (2πfct + φ) where the phase φ is unknown.
The message signal is m (t); it has a bandwidth of W Hz and a
magnitude less than 1 (|m (t) | < 1). The phase φ is unknown. The
instructor drew a diagram (Figure 4.30) for a receiver on the board;
Sammy slept through the description of what the unknown systems
where.
Figure 4.30 Sammy Falls Asleep...
1. What are the signals xc(t) and xs (t)?
2. What would you put in for the unknown systems that would
guarantee that the final output contained the message regardless
of the phase?
HINT: Think of a trigonometric identity that
would prove useful.
3. Sammy may have been asleep, but he can think of a far simpler
receiver. What is it?
188
Problem 4.23: Jamming
Sid Richardson college decides to set up its own AM radio station
KSRR. The resident electrical engineer decides that she can choose
any carrier frequency and message bandwidth for the station. A rival
college decides to jam its transmissions by transmitting a highpower signal that interferes with radios that try to receive KSRR. The
jamming signal jam (t) is what is known as a sawtooth wave
(depicted in Figure 4.31) having a period known to KSRR's engineer.
Figure 4.31 Jamming
Find the spectrum of the jamming signal.
Can KSRR entirely circumvent the attempt to jam it by carefully
choosing its carrier frequency and transmission bandwidth? If so,
find the station's carrier frequency and transmission bandwidth in
terms of T , the period of the jamming signal; if not, show why not.
189
Problem 4.24: AM Stereo
A stereophonic signal consists of a "left" signal l (t) and a "right"
signal r (t) that conveys sounds coming from an orchestra's left and
right sides, respectively. To transmit these two signals
simultaneously, the transmitter first forms the sum signal s+ (t)= l
(t)+ r (t) and the difference signal s− (t)= l (t) − r (t). Then, the
transmitter amplitude-modulates the difference signal with a
sinusoid having frequency 2W , where W is the bandwidth of the left
and right signals. The sum signal and the modulated difference signal
are added, the sum amplitude-modulated to the radio station's
carrier frequency fc, and transmitted. Assume the spectra of the left
and right signals are as shown (Figure 4.32).
Figure 4.32 AM Stereo
1. What is the expression for the transmitted signal? Sketch its
spectrum.
2. Show the block diagram of a stereo AM receiver that can yield the
left and right signals as separate outputs.
3. What signal would be produced by a conventional coherent AM
receiver that expects to receive a standard AM signal conveying a
message signal having bandwidth W ?
190
Problem 4.25: Novel AM Stereo Method
A clever engineer has submitted a patent for a new method for
transmitting two signals simultaneously in the same transmission
bandwidth as commercial AM radio. As shown (Figure 4.33), her
approach is to modulate the positive portion of the carrier with one
signal and the negative portion with a second.
Figure 4.33 Example Transmitter Waveform
In detail the two message signals m1 (t) and m2 (t) are bandlimited to
W Hz and have maximal amplitudes equal to 1. The carrier has a
frequency fc much greater than W . The transmitted signal x (t) is
given by
In all cases, 0 <a< 1. The plot shows the transmitted signal when the
messages are sinusoids: m1 (t)= sin (2πfmt) and m2 (t) = sin (2π2fmt)
where 2fm <W . You, as the patent examiner, must determine
whether the scheme meets its claims and is useful.
191
Problem 4.26: A Radical Radio Idea
An ELEC 241 student has the bright idea of using a square wave
instead of a sinusoid as an AM carrier. The transmitted signal would
have the form
where the message signal m (t) would be amplitude-limited: |m (t) | <
1
1. Assuming the message signal is lowpass and has a bandwidth of W
Hz, what values for the square wave's period T are feasible. In
other words, do some combinations of W and T prevent reception?
2. Assuming reception is possible, can standard radios receive this
innovative AM transmission? If so, show how a coherent receiver
could demodulate it; if not, show how the coherent receiver's
output would be corrupted. Assume that the message bandwidth W
=5 kHz.
Problem 4.27: Secret Communication
An amplitude-modulated secret message m (t) has the following
form.
The message signal has a bandwidth of W Hz and a magnitude less
than 1 (|m (t) | < 1). The idea is to offset the carrier frequency by f0 Hz
from standard radio carrier frequencies. Thus, "of-the-shelf"
coherent demodulators would assume the carrier frequency has fc
Hz. Here, f0 <W .
1. Sketch the spectrum of the demodulated signal produced by a
coherent demodulator tuned to fc Hz.
2. Will this demodulated signal be a "scrambled" version of the
original? If so, how so; if not, why not?
3. Can you develop a receiver that can demodulate the message
without knowing the offset frequency fc?
192
Problem 4.28: Signal Scrambling
An excited inventor announces the discovery of a way of using analog
technology to render music unlistenable without knowing the secret
recovery method. The idea is to modulate the bandlimited message m
(t) by a special periodic signal s (t) that is zero during half of its
period, which renders the message unlistenable and superficially, at
least, unrecoverable (Figure 4.34).
Figure 4.34 Signal Scrambling
1. What is the Fourier series for the periodic signal?
2. What are the restrictions on the period T so that the message
signal can be recovered from m (t) s (t)?
3. ELEC 241 students think they have "broken" the inventor's
scheme and are going to announce it to the world. How would they
recover the original message without having detailed knowledge
of the modulating signal?
4.12 Solutions to Exercises in Chapter 4
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 4.2.1
Because of Euler's relation,
Thus,
, and the other coefficients are zero.
Solution to Exercise 4.2.2
. This quantity clearly corresponds to the periodic pulse signal's average value.
Solution to Exercise 4.3.1
Write the coefficients of the complex Fourier series in Cartesian form as ck = Ak + jBk
and substitute into the expression for the complex Fourier series.
193
Simplifying each term in the sum using Euler’s formula.
We now combine terms that have the same frequency index in magnitude. Because
the signal is real-valued, the coefficients of the complex Fourier series have conjugate
symmetry: c−k = ck*or A−k = Ak and B−k = −Bk. After we add the positive-indexed and
negative-indexed terms, each term in the Fourier series becomes
. To obtain the classic Fourier series (4.11), we must have 2Ak= akand 2Bk = −bk.
Solution to Exercise 4.3.2
The average of a set of numbers is the sum divided by the number of terms. Viewing
signal integration as the limit of a Riemann sum, the integral corresponds to the
average.
Solution to Exercise 4.3.3
We found that the complex Fourier series coefficients are given by
The coefficients are pure imaginary, which means ak =0. The coefficients of the sine
terms are given by bk = − (2Im (ck)) so that
Thus, the Fourier series for the square wave is
Solution to Exercise 4.4.1
The rms value of a sinusoid equals its amplitude divided by
As a half-wave rectified sine wave is zero during half of the period, its rms value is
since the integral of the squared half-wave rectified sine wave equals half that of a
squared sinusoid.
Solution to Exercise 4.4.2
Total harmonic distortion equals
194
Clearly, this quantity is most easily computed in the frequency domain. However, the
numerator equals the square of the signal's rms value minus the power in the average
and the power in the first harmonic.
Solution to Exercise 4.5.1
Total harmonic distortion in the square wave is
Solution to Exercise 4.6.1
N signals directly encoded require a bandwidth of
. Using a binary representation, we need
. For N = 128, the binary-encoding scheme has a factor of
smaller bandwidth. Clearly, binary encoding is superior.
Solution to Exercise 4.6.2
We can use N different amplitude values at only one frequency to represent the
various letters.
Solution to Exercise 4.7.1
Because the filter's gain at zero frequency equals one, the average output values
equal the respective average input values.
Solution to Exercise 4.8.1
Solution to Exercise 4.8.2
. We know that F (S (f)) =
s (−t). Therefore, two Fourier transforms applied to s (t) yields s (−t). We need two
more to get us back where we started.
Solution to Exercise 4.8.3
The signal is the inverse Fourier transform of the triangularly shaped spectrum, and
equals
s\left ( t \right )=W\left ( \frac{sin(\left \pi Wt \right )}{\pi W t} \right )^2
195
Solution to Exercise 4.8.4
The result is most easily found in the spectrum's formula: the power in the signalrelated part of x (t) is half the power of the signal s (t).
Solution to Exercise 4.9.1
The inverse transform of the frequency response is
Multiplying the frequency response by
means subtract from the original signal its time-delayed version. response's timedomain version by Δ results in RC eu (t − Δ). Delaying the frequency resonse’s timedomain version byΔ results in
Subtracting from the undelayed signal
Now we integrate this sum. Because the integral of a sum equals the sum of the
component integrals (integration is linear), we can consider each separately. Because
integration and signal-delay are linear, the integral of a delayed signal equals the
delayed version of the integral. The integral is provided in the example.
Solution to Exercise 4.10.1
If the glottis were linear, a constant input (a zero-frequency sinusoid) should yield a
constant output. The periodic output indicates nonlinear behavior.
Solution to Exercise 4.10.2
In the bottom-left panel, the period is about 0.009 s, which equals a frequency of 111
Hz. The bottom-right panel has a period of about 0.0065 s, a frequency of 154 Hz.
Solution to Exercise 4.10.3
Because males have a lower pitch frequency, the spacing between spectral lines is
smaller. This closer spacing more accurately reveals the formant structure. Doubling
the pitch frequency to 300 Hz for Figure 4.16 (voice spectrum) would amount to
removing every other spectral line.
196
Chapter 5 Digital Signal Processing
5.1 Introduction to Digital Signal Processing
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Not only do we have analog signals -- signals that are real-or complex-valued functions
of a continuous variable such as time or space we can define digital ones as well.
Digital signals are sequences, functions defined only for the integers. We thus use the
notation s (n) to denote a discrete-time one-dimensional signal such as a digital music
recording and s (m, n) for a discrete-"time" two-dimensional signal like a photo taken
with a digital camera. Sequences are fundamentally different than continuous-time
signals. For example, continuity has no meaning for sequences.
Despite such fundamental differences, the theory underlying digital signal processing
mirrors that for analog signals: Fourier transforms, linear filtering, and linear systems
parallel what previous chapters described. These similarities make it easy to
understand the definitions and why we need them, but the similarities should not be
construed as "analog wannabes." We will discover that digital signal processing is not
an approximation to analog processing. We must explicitly worry about the fidelity of
converting analog signals into digital ones. The music stored on CDs, the speech sent
over digital cellular telephones, and the video carried by digital television all evidence
that analog signals can be accurately converted to digital ones and back again.
The key reason why digital signal processing systems have a technological advantage
today is the computer: computations, like the Fourier transform, can be performed
1
quickly enough to be calculated as the signal is produced , and programmability
means that the signal processing system can be easily changed. This flexibility has
obvious appeal, and has been widely accepted in the marketplace. Programmability
means that we can perform signal processing operations impossible with analog
systems (circuits). We will also discover that digital systems enjoy an algorithmic
advantage that contributes to rapid processing speeds: Computations can be
restructured in non-obvious ways to speed the processing. This flexibility comes at a
price, a consequence of how computers work. How do computers perform signal
processing?
1. Taking a systems viewpoint for the moment, a system that produces its output as rapidly as the input arises is said to be
a real-time system. All analog systems operate in real time; digital ones that depend on a computer to perform system
computations may or may not work in real time. Clearly, we need real-time signal processing systems. Only recently
have computers become fast enough to meet real-time requirements while performing non-trivial signal processing.
197
5.2 Introduction to Computer Organization
5.2.1 Computer Architecture
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
To understand digital signal processing systems, we must understand a little about
how computers compute. The modern definition of a computer is an electronic
device that performs calculations on data, presenting the results to humans or other
computers in a variety of (hopefully useful) ways.
Organization of a Simple Computer
Figure 5.1 Generic computer hardware organization.
The generic computer contains input devices (keyboard, mouse, A/D (analog-todigital) converter, etc.), a computational unit, and output devices (monitors, printers,
D/A converters). The computational unit is the computer's heart, and usually consists
of a central processing unit (CPU), a memory, and an input/output (I/O) interface.
What I/O devices might be present on a given computer vary greatly.
• A simple computer operates fundamentally in discrete time. Computers are
clocked devices, in which computational steps occur periodically according to
ticks of a clock. This description be lies clock speed: When you say "I have a 1 GHz
computer," you mean that your computer takes 1 nanosecond to perform each
step. That is incredibly fast! A "step" does not, unfortunately, necessarily mean a
computation like an addition; computers break such computations down into
several stages, which means that the clock speed need not express the
computational speed. Computational speed is expressed in units of millions of
instructions/second (Mips). Your 1 GHz computer (clock speed) may have a
computational speed of 200 Mips.
• Computers perform integer (discrete-valued) computations. Computer
calculations can be numeric (obeying the laws of arithmetic), logical (obeying the
laws of an algebra), or symbolic (obeying any law you like).4 Each computer
instruction that performs an elementary numeric calculation an addition, a
multiplication, or a division does so only for integers. The sum or product of two
198
integers is also an integer, but the quotient of two integers is likely to not be an
integer. How does a computer deal with numbers that have digits to the right of
the decimal point? This problem is addressed by using the so-called floatingpoint representation of real numbers. At its heart, however, this representation
relies on integer-valued computations.
5.2.2 Representing Numbers
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Focusing on numbers, all numbers can represented by the positional notation
system. The b-ary positional representation system uses the position of digits ranging
from 0 to b-1 to denote a number. The quantity b is known as the base of the number
system. Mathematically, positional systems represent the positive integer n as
and we succinctly express n in base-b as nb = dN dN−1 ...d0. The number 25 in base 10
1
0
equals 2×10 +5×10 , so that the digits representing this number are d0 =5, d1 =2, and
4
all other dk equal zero. This same number in binary (base 2) equals 11001 (1 × 2 +1 ×
3
2
1
0
2 +0 × 2 +0 × 2 +1 × 2 ) and 19 in hexadecimal (base 16). Fractions between zero
and one are represented the same way.
All numbers can be represented by their sign, integer and fractional parts. Complex
numbers (Section 2.1) can be thought of as two real numbers that obey special rules
to manipulate them. Humans use base 10, commonly assumed to be due to us having
ten fingers. Digital computers use the base 2 or binary number representation, each
digit of which is known as a bit (binary digit).
Figure 5.2 Number representations on computers The various ways numbers are represented in binary
are illustrated. The number of bytes for the exponent and mantissa components of foating point numbers
varies.
Here, each bit is represented as a voltage that is either "high" or "low," thereby
representing "1" or "0," respectively. To represent signed values, we tack on a special
bit the sign bit to express the sign. The computer's memory consists of an ordered
199
sequence of bytes, a collection of eight bits. A byte can therefore represent an
unsigned number ranging from 0 to 255. If we take one of the bits and make it the
sign bit, we can make the same byte to represent numbers ranging from −128 to 127.
But a computer cannot represent all possible real numbers. The fault is not with the
binary number system; rather having only a finite number of bytes is the problem.
While a gigabyte of memory may seem to be a lot, it takes an infinite number of bits to
represent π. Since we want to store many numbers in a computer's memory, we are
restricted to those that have a finite binary representation. Large integers can be
represented by an ordered sequence of bytes. Common lengths, usually expressed in
terms of the number of bits, are 16, 32, and 64. Thus, an unsigned 32-bit number can
32
represent integers ranging between 0 and 2 − 1 (4,294,967,295), a number almost
big enough to enumerate every human in the world!6
Exercise 5.2.1
For both 32-bit and 64-bit integer representations, what are the
largest numbers that can be represented if a sign bit must also be
included.
While this system represents integers well, how about numbers having nonzero digits
to the right of the decimal point? In other words, how are numbers that have
fractional parts represented? For such numbers, the binary representation system is
used, but with a little more complexity. The floating-point system uses a number of
bytes -typically 4 or 8 -to represent the number, but with one byte (sometimes two
bytes) reserved to represent the exponent e of a power-of-two multiplier for the
number -the mantissa m -expressed by the remaining bytes.
The mantissa is usually taken to be a binary fraction having a magnitude in the range
2
, which means that the binary representation is such that d−1 =1. The number zero is
an exception to this rule, and it is the only floating point number having a zero
fraction. The sign of the mantissa represents the sign of the number and the exponent
can be a signed integer.
A computer's representation of integers is either perfect or only approximate, the
latter situation occurring when the integer exceeds the range of numbers that a
limited set of bytes can represent. Floating point representations have similar
representation problems: if the number x can be multiplied/divided by enough
powers of two to yield a fraction lying between 1/2 and 1 that has a finite binaryfraction representation, the number is represented exactly in foating point. Otherwise,
we can only represent the number approximately, not catastrophically in error as with
integers. For example, the number 2.5 equals 0.625 × 22, the fractional part of which
3
has an exact binary representation. However, the number 2.6 does not have an
exact binary representation, and only be represented approximately in floating point.
2. In some computers, this normalization is taken to an extreme: the leading binary digit is not explicitly expressed,
providing an extra bit to represent the mantissa a little more accurately. This convention is known as the hidden-ones
notation.
3. See if you can nd this representation.
200
In single precision floating point numbers, which require 32 bits (one byte for the
exponent and the remaining 24 bits for the mantissa), the number 2.6 will be
represented as 2.600000079.... Note that this approximation has a much longer
decimal expansion. This level of accuracy may not suffice in numerical calculations.
Double precision floatingpoint numbers consume 8 bytes, and quadruple
precision 16 bytes. The more bits used in the mantissa, the greater the accuracy. This
increasing accuracy means that more numbers can be represented exactly, but there
are always some that cannot. Such inexact numbers have an infinite binary
4
representation. Realizing that real numbers can be only represented approximately
is quite important, and underlies the entire field of numerical analysis, which seeks to
predict the numerical accuracy of any computation.
Exercise 5.2.2
What are the largest and smallest numbers that can be represented in
32-bit foating point? in 64-bit foating point that has sixteen bits
allocated to the exponent? Note that both exponent and mantissa
require a sign bit.
So long as the integers aren't too large, they can be represented exactly in a computer
using the binary positional notation. Electronic circuits that make up the physical
computer can add and subtract integers without error. (This statement isn't quite true;
when does addition cause problems?)
5.2.3 Computer Arithmetic and Logic
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The binary addition and multiplication tables are
4. Note that there will always be numbers that have an innite representation in any chosen positional system. The choice
of base denes which do and which don't. If you were thinking that base 10 numbers would solve this inaccuracy, note
that 1=3 = 0:333333:::: has an innite representation in decimal (and binary for that matter), but has nite representation in
base 3.
201
Note that if carries are ignored, subtraction of two single-digit binary numbers yields
the same bit as addition. Computers use high and low voltage values to express a bit,
and an array of such voltages express numbers akin to positional notation. Logic
circuits perform arithmetic operations.
Exercise 5.2.3
Add twenty-five and seven in base 2. Note the carries that might
occur. Why is the result "nice"? The variables of logic indicate truth
or falsehood. A ∩ B, the AND of A and B, represents a statement that
both A and B must be true for the statement to be true. You use this
kind of statement to tell search engines that you want to restrict hits
to cases where both of the events A and B occur. A ∪ B, the OR of A
and B, yields a value of truth if either is true. Note that if we
represent truth by a "1" and falsehood by a "0," binary
multiplication corresponds to AND and addition (ignoring carries)
to XOR. XOR, the exclusive or operator, equals the union of A ∪ B and
A ∩ B. The Irish mathematician George Boole discovered this
equivalence in the mid-nineteenth century. It laid the foundation for
what we now call Boolean algebra, which expresses as equations
logical statements. More importantly, any computer using base-2
representations and arithmetic can also easily evaluate logical
statements. This fact makes an integer-based computational device
much more powerful than might be apparent.
5.3 The Sampling Theorem
5.3.1 Analog-to-Digital Conversion
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Because of the way computers are organized, signal must be represented by a fnite
number of bytes. This restriction means that both the time axis and the amplitude
axis must be quantized: They must each be a multiple of the integers. l2 Quite
surprisingly, the Sampling Theorem allows us to quantize the time axis without error
for some signals. The signals that can be sampled without introducing error are
interesting, and as described in the next section, we can make a signal "samplable" by
filtering. In contrast, no one has found a way of performing the amplitude
quantization step without introducing an unrecoverable error. Thus, a signal's value
can no longer be any real number. Signals processed by digital computers must be
discrete-valued: their values must be proportional to the integers. Consequently,
analog-to-digital conversion introduces error.
202
5.4 The Sampling Theorem
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Digital transmission of information and digital signal processing all require signals to
first be "acquired" by a computer. One of the most amazing and useful results in
electrical engineering is that signals can be converted from a function of time into a
sequence of numbers without error: We can convert the numbers back into the
signal with (theoretically) no error. Harold Nyquist, a Bell Laboratories engineer, first
derived this result, known as the Sampling Theorem, in the 1920s. It found no real
application back then. Claude Shannon, also at Bell Laboratories, revived the result
once computers were made public after World War II.
The sampled version of the analog signal s (t) is s (nTs), with Ts known as the sampling
interval. Clearly, the value of the original signal at the sampling times is preserved;
the issue is how the signal values between the samples can be reconstructed since
they are lost in the sampling process. To characterize sampling, we approximate it as
the product x (t)= s (t) PTs (t), with PTs (t) being the periodic pulse signal. The resulting
signal, as shown in Figure 5.3 (Sampled Signal), has nonzero values only during the
time intervals
Figure 5.3 Sampled Signal The waveform of an example signal is shown in the top plot and its sampled
version in the bottom.
For our purposes here, we center the periodic pulse signal about the origin so that its
Fourier series coefficients are real (the signal is even).
203
If the properties of s (t) and the periodic pulse signal are chosen properly, we can
recover s (t) from x (t) by filtering.
To understand how signal values between the samples can be "filled" in, we need to
calculate the sampled signal's spectrum. Using the Fourier series representation of the
periodic sampling signal,
Considering each term in the sum separately, we need to know the spectrum of the
product of the complex exponential and the signal. Evaluating this transform directly
is quite easy.
Thus, the spectrum of the sampled signal consists of weighted (by the coefficients ck)
and delayed versions of the signal's spectrum (Figure 5.4 (aliasing)).
In general, the terms in this sum overlap each other in the frequency domain,
rendering recovery of the original signal impossible. This unpleasant phenomenon is
known as aliasing.
Figure 5.4 aliasing The spectrum of some bandlimited (to W Hz) signal is shown in the top plot. If the
sampling interval Ts is chosen too large relative to the bandwidth W, aliasing will occur. In the bottom plot,
the sampling interval is chosen sufficiently small to avoid aliasing. Note that if the signal were not
bandlimited, the component spectra would always overlap.
If, however, we satisfy two conditions:
204
• The signal s (t) is bandlimited has power in a restricted frequency range to W Hz,
and
• the sampling interval Ts is small enough so that the individual components in the
sum do not overlap Ts < 1/2W ,
aliasing will not occur. In this delightful case, we can recover the original signal by
lowpass filtering x (t) with a filter having a cutoff frequency equal to W Hz. These two
conditions ensure the ability to recover a bandlimited signal from its sampled version:
We thus have the Sampling Theorem.
Exercise 5.3.1
The Sampling Theorem (as stated) does not mention the pulse width
Δ. What is the effect of this parameter on our ability to recover a
signal from its samples (assuming the Sampling Theorem's two
conditions are met)?
The frequency
, known today as the Nyquist frequency and the Shannon sampling frequency,
corresponds to the highest frequency at which a signal can contain energy and remain
compatible with the Sampling Theorem. High-quality sampling systems ensure that no
aliasing occurs by unceremoniously lowpass filtering the signal (cutoff frequency being
slightly lower than the Nyquist frequency) before sampling. Such systems therefore
vary the anti-aliasing filter’s cutoff frequency as the sampling rate varies. Because
such quality features cost money, many sound cards do not have anti-aliasing filters
or, for that matter, post-sampling filters. They sample at high frequencies, 44.1 kHz for
example, and hope the signal contains no frequencies above the Nyquist frequency
(22.05 kHz in our example). If, however, the signal contains frequencies beyond the
sound card's Nyquist frequency, the resulting aliasing can be impossible to remove.
Exercise 5.3.2
To gain a better appreciation of aliasing, sketch the spectrum of a
sampled square wave. For simplicity consider only the spectral
repetitions centered at
Let the sampling interval Ts be 1; consider two values for the square
wave's period: 3.5 and 4. Note in particular where the spectral lines
go as the period decreases; some will move to the left and some to
the right. What property characterizes the ones going the same
direction?
If we satisfy the Sampling Theorem's conditions, the signal will change only slightly
during each pulse. As we narrow the pulse, making Δ smaller and smaller, the nonzero
values of the signal s(t) pTs (t) will simply be s (nTs), the signal's samples. If indeed the
Nyquist frequency equals the signal's highest frequency, at least two samples will
205
occur within the period of the signal's highest frequency sinusoid. In these ways, the
sampling signal captures the sampled signal's temporal variations in a way that leaves
all the original signal's structure intact.
Exercise 5.3.3
What is the simplest bandlimited signal? Using this signal, convince
yourself that less than two samples/period will not suffice to specify
it. If the sampling rate
is not high enough, what signal would your resulting undersampled
signal become?
5.5 Amplitude Quantization
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The Sampling Theorem says that if we sample a bandlimited signal s (t) fast enough, it
can be recovered without error from its samples s (nTs), n ∈{..., −1, 0, 1,... }. Sampling
is only the first phase of acquiring data into a computer: Computational processing
further requires that the samples be quantized: analog values are converted into
digital (Section 1.2.2: Digital Signals) form. In short, we will have performed analog-todigital (A/D) conversion.
206
Figure 5.5 A three-bit A/D converter A three-bit A/D converter assigns voltage in the range [−1, 1] to one
of eight integers between 0 and 7. For example, all inputs having values lying between 0.5 and 0.75 are
assigned the integer value six and, upon conversion back to an analog value, they all become 0.625. The
width of a single quantization interval Δ equals . The bottom panel shows a signal going through the analogto-digital converter, where B is the number of bits used in the A/D conversion process (3 in the case
depicted here). First it is sampled, then amplitude-quantized to three bits. Note how the sampled signal
waveform becomes distorted after amplitude quantization. For example the two signal values between 0.5
and 0.75 become 0.625. This distortion is irreversible; it can be reduced (but not eliminated) by using more
bits in the A/D converter.
A phenomenon reminiscent of the errors incurred in representing numbers on a
computer prevents signal amplitudes from being converted with no error into a binary
number representation. In analog-to-digital conversion, the signal is assumed to lie
within a predefined range. Assuming we can scale the signal without affecting the
information it expresses, we'll define this range to be [−1, 1]. Furthermore, the A/D
converter assigns amplitude values in this range to a set of integers. A B-bit converter
produces one of the integers {0, 1,..., 2B– 1}for each sampled input. Figure 5.5 shows
how a three-bit A/D converter assigns input values to the integers. We define a
quantization interval to be the range of values assigned to the same integer. Thus,
for our example three-bit A/D converter, the quantization interval Δ is 0.25; in general,
it is
.
Exercise 5.4.1
Recalling the plot of average daily highs in this frequency domain
problem (Problem 4.5), why is this plot so jagged? Interpret this
effect in terms of analog-to-digital conversion.
207
Because values lying anywhere within a quantization interval are assigned the same
value for computer processing, the original amplitude value cannot be recovered
without error. Typically, the D/A converter, the device that converts integers to
amplitudes, assigns an amplitude equal to the value lying halfway in the quantization
interval. The integer 6 would be assigned to the amplitude 0.625 in this scheme. The
error introduced by converting a signal from analog to digital form by sampling and
amplitude quantization then back again would be half the quantization interval for
each amplitude value. Thus, the so-called A/D error equals half the width of a
quantization interval:
. As we have fixed the input-amplitude range, the more bits available in the A/D
converter, the smaller the quantization error.
To analyze the amplitude quantization error more deeply, we need to compute the
signal-to-noise ratio, which equals the ratio of the signal power and the quantization
error power. Assuming the signal is a sinusoid, the signal power is the square of the
rms amplitude: power
The illustration (Figure 5.6) details a single quantization interval.
Figure 5.6 A single quantization interval A single quantization interval is shown, along with a typical
signal's value before amplitude quantization s (nTs) and after Q (s (nTs)). E denotes the error thus incurred.
Its width is Δ and the quantization error is denoted by E. To find the power in the
quantization error, we note that no matter into which quantization interval the signal's
value falls, the error will have the same characteristics. To calculate the rms value, we
must square the error and average it over the interval.
Since the quantization interval width for a B-bit converter equals
we find that the signal-tonoise ratio for the analog-to-digital conversion process
equals
208
Thus, every bit increase in the A/D converter yields a 6 dB increase in the signal-tonoise ratio. The constant term 10log1.5 equals 1.76.
Exercise 5.4.2
This derivation assumed the signal's amplitude lay in the range [−1,
1]. What would the amplitude quantization signal-to-noise ratio be if
it lay in the range [−A, A]?
Exercise 5.4.3
How many bits would be required in the A/D converter to ensure that
the maximum amplitude quantization error was less than 60 db
smaller than the signal's peak value?
Exercise 5.4.4
Music on a CD is stored to 16-bit accuracy. To what signal-to-noise
ratio does this correspond?
Once we have acquired signals with an A/D converter, we can process them using
digital hardware or software. It can be shown that if the computer processing is linear,
the result of sampling, computer processing, and unsampling is equivalent to some
analog linear system. Why go to all the bother if the same function can be
accomplished using analog techniques? Knowing when digital processing excels and
when it does not is an important issue.
5.6 Discrete-Time Signals and Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Mathematically, analog signals are functions having as their independent variables
continuous quantities, such as space and time. Discrete-time signals are functions
defined on the integers; they are sequences. As with analog signals, we seek ways of
decomposing discrete-time signals into simpler components. Because this approach
leads to a better understanding of signal structure, we can exploit that structure to
represent information (create ways of representing information with signals) and to
extract information (retrieve the information thus represented). For symbolic-valued
signals, the approach is different: We develop a common representation of all
symbolic-valued signals so that we can embody the information they contain in a
unifed way. From an information representation perspective, the most important
issue becomes, for both real-valued and symbolic-valued signals, efficiency: what is
the most parsimonious and compact way to represent information so that it can be
extracted later.
209
5.6.1 Real-and Complex-valued Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A discrete-time signal is represented symbolically as s (n), where n = {..., −1, 0, 1,... }.
Figure 5.7 Cosine The discrete-time cosine signal is plotted as a stem plot. Can you find the formula for this
signal?
We usually draw discrete-time signals as stem plots to emphasize the fact they are
functions defined only on the integers. We can delay a discrete-time signal by an
integer just as with analog ones. A signal delayed by m samples has the expression s
(n − m).
5.6.2 Complex Exponentials
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The most important signal is, of course, the complex exponential sequence.
Note that the frequency variable f is dimensionless and that adding an integer to the
frequency of the discrete-time complex exponential has no effect on the signal's value.
This derivation follows because the complex exponential evaluated at an integer
multiple of 2π equals one. Thus, we need only consider frequency to have a value in
some unit-length interval.
5.6.3 Sinusoids
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Discrete-time sinusoids have the obvious form s (n)= Acos (2πfn + Ï•). As opposed to
analog complex exponentials and sinusoids that can have their frequencies be any
real value, frequencies of their discrete-time counterparts yield unique waveforms
only when f lies in the interval
210
This choice of frequency interval is arbitrary; we can also choose the frequency to lie
in the interval [0,1). How to choose a unit-length interval for a sinusoid's frequency will
become evident later.
5.6.4 Unit Sample
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The second-most important discrete-time signal is the unit sample, which is defined
to be
(5.14)
Figure 5.8 The unit sample.
Examination of a discrete-time signal's plot, like that of the cosine signal shown in
Figure 5.7(Cosine), reveals that all signals consist of a sequence of delayed and scaled
unit samples. Because the value of a sequence at each integer m is denoted by s(m)
and the unit sample delayed to occur at m is written δ(n - m), we can decompose any
signal as a sum of unit samples delayed to the appropriate location and scaled by the
signal value.
This kind of decomposition is unique to discrete-time signals, and will prove useful
subsequently.
5.6.5 Unit Step
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The unit step in discrete-time is well-defined at the origin, as opposed to the situation
with analog signals.
(5.16)
211
5.6.6 Symbolic Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
An interesting aspect of discrete-time signals is that their values do not need to be real
numbers. We do have real-valued discrete-time signals like the sinusoid, but we also
have signals that denote the sequence of characters typed on the keyboard. Such
characters certainly aren't real numbers, and as a collection of possible signal values,
they have little mathematical structure other than that they are members of a set.
More formally, each element of the symbolic-valued signal s(n) takes on one of the
values {a1,...,aK} which comprise the alphabet A. This technical terminology does not
mean we restrict symbols to being members of the English or Greek alphabet. They
could represent keyboard characters, bytes (8-bit quantities), integers that convey
daily temperature. Whether controlled by software or not, discrete-time systems are
ultimately constructed from digital circuits, which consist entirely of analog circuit
elements. Furthermore, the transmission and reception of discrete-time signals, like email, is accomplished with analog signals and systems. Understanding how discretetime and analog signals and systems intertwine is perhaps the main goal of this
course.
5.6.7 Discrete-Time Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Discrete-time systems can act on discrete-time signals in ways similar to those found
in analog signals and systems. Because of the role of software in discrete-time
systems, many more different systems can be envisioned and "constructed" with
programs than can be with analog signals. In fact, a special class of analog signals can
be converted into discrete-time signals, processed with software, and converted back
into an analog signal, all without the incursion of error. For such signals, systems can
be easily produced in software, with equivalent analog realizations difcult, if not
impossible, to design.
5.7 Discrete-Time Fourier Transform (DTFT)
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The Fourier transform of the discrete-time signal s (n) is defined to be
Frequency here has no units. As should be expected, this Definition is linear, with the
transform of a sum of signals equaling the sum of their transforms. Real-valued
signals have conjugate-symmetric spectra:
212
Exercise 5.6.1
A special property of the discrete-time Fourier transform is that it is periodic with
period one:
Derive this property from the Definition of the DTFT.
Because of this periodicity, we need only plot the spectrum over one period to
understand completely the spectrum's structure; typically, we plot the spectrum over
the frequency range
When the signal is real-valued, we can further simplify our plotting chores by showing
the spectrum only over
the spectrum at negative frequencies can be derived from positive-frequency spectral
values. When we obtain the discrete-time signal via sampling an analog signal, the
Nyquist frequency (p. 176) corresponds to the discrete-time frequency
To show this, note that a sinusoid having a frequency equal to the Nyquist frequency
has a sampled waveform that equals
The exponential in the DTFT at frequency
equals
meaning that discrete-time frequency equals analog frequency multiplied by the
sampling interval
fD = fATs
(5.18)
fD and fA represent discrete-time and analog frequency variables, respectively. The
aliasing figure (Figure 5.4) provides another way of deriving this result. As the duration
of each pulse in the periodic sampling signal PTs(t) narrows, the amplitudes of the
signal's spectral repetitions, which are governed by the Fourier series coefficients
(4.10) of PTs(t), become increasingly equal. Examination of the periodic pulse signal
(Figure 4.1) reveals that as Δ decreases, the value of c0, the largest Fourier coefficient,
decreases to zero:
Thus, to maintain a mathematically viable Sampling Theorem, the amplitude A must
increase as
213
becoming infinitely large as the pulse duration decreases. Practical systems use a
small value of Δ, say 0.1 · Ts and use amplifiers to rescale the signal. Thus, the sampled
signal's spectrum becomes periodic with period
. Thus, the Nyquist frequency
corresponds to the frequency
Example 5.1
Let's compute the discrete-time Fourier transform of the exponentially decaying
n
sequence s(n)=a u(n), where u (n) is the unit-step sequence. Simply plugging the signal's
expression into the Fourier transform formula,
This sum is a special case of the geometric series.
Thus, as long as |a| < 1, we have our Fourier transform.
Using Euler's relation, we can express the magnitude and phase of this spectrum.
No matter what value of a we choose, the above formulae clearly demonstrate the
periodic nature of the spectra of discrete-time signals. Figure 5.9 (Spectrum of
exponential signal) shows indeed that the spectrum is a periodic function. We need
only consider the spectrum between
and
214
to unambiguously define it. When a> 0, we have a lowpass spectrum the spectrum
diminishes as frequency increases from 0 to
with increasing a leading to a greater low frequency content; for a< 0, we have a
highpass spectrum (Figure 5.10 (Spectra of exponential signals)).
Figure 5.9 Spectrum of exponential signall
The spectrum of the exponential signal (a = 0:5) is shown over the frequency range
[-2,2], clearly demonstrating the periodicity of all discrete-time spectra. The angle has
units of degrees.
Figure 5.10 Spectra of exponential signals
The spectra of several exponential signals are shown. What is the apparent
relationship between the spectra for a = 0.5 and a = -0.5?
Example 5.2
215
Analogous to the analog pulse signal, let's find the spectrum of the length-N pulse
sequence.
(5.24)
The Fourier transform of this sequence has the form of a truncated geometric series.
(5.25)
For the so-called finite geometric series, we know that
(5.26)
for all values of α.
Exercise 5.6.2
Derive this formula for the finite geometric series sum. The "trick" is to consider the
difference between the series' sum and the sum of the series multiplied by α. Applying
this result yields (Figure 5.11 (Spectrum of length-ten pulse).)
The ratio of sine functions has the generic form of
which is known as the discrete-time sinc function dsinc (x). Thus, our transform can
be concisely expressed as
The discrete-time pulse's spectrum contains many ripples, the number of which
increase with N, the pulse's duration.
216
Figure 5.11 Spectrum of length-ten pulse
The spectrum of a length-ten pulse is shown. Can you explain the rather complicated
appearance of the phase?
The inverse discrete-time Fourier transform is easily derived from the following
relationship:
Therefore, we find that
The Fourier transform pairs in discrete-time are
The properties of the discrete-time Fourier transform mirror those of the analog
Fourier transform. The DTFT properties table ("Properties of the DTFT"
<http://legacy.cnx.org/content/m0506/latest/>) shows similarities and diferences. One
important common property is Parseval's Theorem.
To show this important property, we simply substitute the Fourier transform
expression into the frequency-domain expression for power.
217
Using the orthogonality relation (5.28), the integral equals δ (m − n), where δ (n) is the
unit sample (Figure 5.8: Unit sample). Thus, the double sum collapses into a single
sum because nonzero values occur only when n = m, giving Parseval's Theorem as a
result. We term
the energy in the discrete-time signal s(n) in spite of the fact that discrete-time signals
don't consume (or produce for that matter) energy. This terminology is a carry-over
from the analog world.
Exercise 5.6.3
Suppose we obtained our discrete-time signal from values of the product s(t)pTs(t),
where the duration of the component pulses in pTs(t) is Δ. How is the discrete-time
signal energy related to the total energy contained in s(t)? Assume the signal is
bandlimited and that the sampling rate was chosen appropriate to the Sampling
Theorem's conditions.
5.8 Discrete Fourier Transforms (DFT)
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The discrete-time Fourier transform (and the continuous-time transform as well) can
be evaluated when we have an analytic expression for the signal. Suppose we just
have a signal, such as the speech signal used in the previous chapter, for which there
is no formula. How then would you compute the spectrum? For example, how did we
compute a spectrogram such as the one shown in the speech signal example (Figure
4.17: spectrogram)? The Discrete Fourier Transform (DFT) allows the computation of
spectra from discrete-time data. While in discrete-time we can exactly calculate
spectra, for analog signals no similar exact spectrum computation exists. For analogsignal spectra, use must build special devices, which turn out in most cases to consist
of A/D converters and discrete-time computations. Certainly discrete-time spectral
analysis is more flexible than continuous-time spectral analysis.
The formula for the DTFT (5.17) is a sum, which conceptually can be easily computed
save for two issues.
• Signal duration. The sum extends over the signal's duration, which must be fnite
to compute the signal's spectrum. It is exceedingly difcult to store an infnitelength signal in any case, so we'll assume that the signal extends over [0,N − 1].
• Continuous frequency. Subtler than the signal duration issue is the fact that the
frequency variable is continuous: It may only need to span one period, like
218
or [0, 1], but the DTFT formula as it stands requires evaluating the spectra at
all frequencies within a period. Let's compute the spectrum at a few frequencies;
the most obvious ones are the equally spaced ones
We thus define the discrete Fourier transform (DFT) to be
Here, S(k) is shorthand for
We can compute the spectrum at as many equally spaced frequencies as we like. Note
that you can think about this computationally motivated choice as sampling the
spectrum; more about this interpretation later. The issue now is how many
frequencies are enough to capture how the spectrum changes with frequency. One
way of answering this question is determining an inverse discrete Fourier transform
formula: given S(k), k = {0,...,K − 1} how do we find s(n), n = {0,...,N − 1}? Presumably, the
formula will be of the form
Substituting the DFT formula in this prototype inverse transform yields
Note that the orthogonality relation we use so often has a different character now.
We obtain nonzero value whenever the two indices difer by multiples of K. We can
express this result as
Thus, our formula becomes
The integers n and m both range over {0,...,N − 1}. To have an inverse transform, we
need the sum to be a single unit sample for m, n in this range. If it did not, then s (n)
would equal a sum of values, and we would not have a valid transform: Once going
into the frequency domain, we could not get back unambiguously! Clearly, the term
l =0 always provides a unit sample (we'll take care of the factor of K soon). If we
219
evaluate the spectrum at fewer frequencies than the signal's duration, the term
corresponding to m = n + K will also appear for some values of m, n = {0,...,N - 1}. This
situation means that our prototype transform equals s (n) + s (n + K) for some values of
n. The only way to eliminate this problem is to require K ≥ N: We must have at least as
many frequency samples as the signal's duration. In this way, we can return from the
frequency domain we entered via the DFT.
Exercise 5.7.1
When we have fewer frequency samples than the signal's duration, some discretetime signal values equal the sum of the original signal values. Given the sampling
interpretation of the spectrum, characterize this effect a different way.
Another way to understand this requirement is to use the theory of linear equations. If
we write out the expression for the DFT as a set of linear equations,
.....
we have K equations in N unknowns if we want to find the signal from its sampled
spectrum. This require ment is impossible to fulfll if K < N; we must have K ≥ N. Our
orthogonality relation essentially says that if we have a sufcient number of equations
(frequency samples), the resulting set of equations can indeed be solved.
By convention, the number of DFT frequency values K is chosen to equal the signal's
duration N. The discrete Fourier transform pair consists of
Discrete Fourier Transform Pair
Example 5.3
Use this demonstration to perform DFT analysis of a signal.
This media object is a LabVIEW VI. Please view or download it at
<DFTanalysis.llb>
Example 5.4
Use this demonstration to synthesize a signal from a DFT sequence.
This media object is a LabVIEW VI. Please view or download it at
<DFT_Component_Manipulation.llb>
220
5.9 DFT: Computational Complexity
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We now have a way of computing the spectrum for an arbitrary signal: The Discrete
Fourier Transform (DFT) (5.33) computes the spectrum at N equally spaced
frequencies from a length-N sequence. An issue that never arises in analog
"computation," like that performed by a circuit, is how much work it takes to perform
the signal processing operation such as filtering. In computation, this consideration
translates to the number of basic computational steps required to perform the
needed processing. The number of steps, known as the complexity, becomes
equivalent to how long the computation takes (how long must we wait for an answer).
Complexity is not so much tied to specific computers or programming languages but
to how many steps are required on any computer. Thus, a procedure's stated
complexity says that the time taken will be proportional to some function of the
amount of data used in the computation and the amount demanded.
For example, consider the formula for the discrete Fourier transform. For each
frequency we choose, we must multiply each signal value by a complex number and
add together the results. For a real-valued signal, each real-times-complex
multiplication requires two real multiplications, meaning we have 2N multiplications to
perform. To add the results together, we must keep the real and imaginary parts
separate. Adding N numbers requires N - 1 additions. Consequently, each frequency
requires 2N + 2(N - 1) = 4N - 2 basic computational steps. As we have N frequencies,
the total number of computations is N(4N - 2).
In complexity calculations, we only worry about what happens as the data lengths
2
increase, and take the dominant term here the 4N term as reflecting how much work
is involved in making the computation. As multiplicative constants don't matter since
2
we are making a "proportional to" evaluation, we find the DFT is an O(N )
computational procedure. This notation is read "order N-squared". Thus, if we double
the length of the data, we would expect that the computation time to approximately
quadruple.
Exercise 5.8.1
In making the complexity evaluation for the DFT, we assumed the data to be real.
Three questions emerge. First of all, the spectra of such signals have conjugate
symmetry, meaning that negative frequency components (
in the DFT (5.33)) can be computed from the corresponding positive frequency
components. Does this symmetry change the DFT's complexity? Secondly, suppose the
data are complex-valued; what is the DFT's complexity now? Finally, a less important
but interesting question is suppose we want K frequency values instead of N; now
what is the complexity?
221
5.10 Fast Fourier Transform (FFT)
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
One wonders if the DFT can be computed faster: Does another computational
procedure an algorithm exist that can compute the same quantity, but more
efficiently. We could seek methods that reduce the constant of proportionality, but do
2
not change the DFT's complexity O(N ). Here, we have something more dramatic in
mind: Can the computations be restructured so that a smaller complexity results?
In 1965, IBM researcher Jim Cooley and Princeton faculty member John Tukey
developed what is now known as the Fast Fourier Transform (FFT). It is an algorithm
for computing that DFT that has order O(NlogN) for certain length inputs. Now when
the length of data doubles, the spectral computational time will not quadruple as with
the DFT algorithm; instead, it approximately doubles. Later research showed that no
algorithm for computing the DFT could have a smaller complexity than the FFT.
Surprisingly, historical work has shown that Gauss2l in the early nineteenth century
developed the same algorithm, but did not publish it! After the FFT's rediscovery, not
only was the computation of a signal's spectrum greatly speeded, but also the added
feature of algorithm meant that computations had flexibility not available to analog
implementations.
Exercise 5.9.1
Before developing the FFT, let's try to appreciate the algorithm's impact. Suppose a
short-length transform takes 1 ms. We want to calculate a transform of a signal that is
10 times longer. Compare how much longer a straightforward implementation of the
DFT would take in comparison to an FFT, both of which compute exactly the same
quantity.
L
To derive the FFT, we assume that the signal's duration is a power of two: N=2 .
Consider what happens to the even-numbered and odd-numbered elements of the
sequence in the DFT calculation.
(5.39)
Each term in square brackets has the form of a
222
DFT. The first one is a DFT of the even-numbered elements, and the second of the
odd-numbered elements. The first DFT is combined with the second multiplied by the
complex exponential
The half-length transforms are each evaluated at frequency indices k = 0, ..., N - 1.
Normally, the number of frequency indices in a DFT calculation range between zero
and the transform length minus one. The computational advantage of the FFT
comes from recognizing the periodic nature of the discrete Fourier transform. The FFT
simply reuses the computations made in the half-length transforms and combines
them through additions and the multiplication by e, which is not periodic over
Figure 5.12 (Length-8 DFT decomposition) illustrates this decomposition. As it stands,
we now compute
transforms
multiply one of them by the complex exponential (complexity O(N)), and add the
results (complexity O(N)). At this point, the total complexity is still dominated by the
half-length DFT calculations, but the proportionality coefficient has been reduced.
L
Now for the fun. Because N=2 , each of the half-length transforms can be reduced to
two quarter-length transforms, each of these to two eighth-length ones, etc. This
decomposition continues until we are left with length-2 transforms. This transform is
quite simple, involving only additions. Thus, the first stage of the FFT has
length-2 transforms (see the bottom part of Figure 5.12 (Length-8 DFT
decomposition)). Pairs of these transforms are combined by adding one to the other
multiplied by a complex exponential. Each pair requires 4 additions and 2
multiplications, giving a total number of computations equaling
This number of computations does not change from stage to stage. Because the
number of stages, the number of times the length can be divided by two, equals
log2N, the number of arithmetic operations equals
which makes the complexity of the FFT O(Nlog2N).
223
Figure 5.12 Length-8 DFT decomposition
Doing an example will make computational savings more obvious. Let's look at the
details of a length-8 DFT. As shown on Figure 5.13 (Butterfy), we first decompose the
DFT into two length-4 DFTs, with the outputs added and subtracted together in pairs.
Considering Figure 5.13 (Butterfy) as the frequency index goes from 0 through 7, we
recycle values from the length-4 DFTs into the fnal calculation because of the
periodicity of the DFT output. Examining how pairs of outputs are collected together,
we create the basic computational element known as a butterfy (Figure 5.13
(Butterfy)).
Figure 5.13 Butterfy
The basic computational element of the fast Fourier transform is the butterfly. It takes two complex
numbers, represented by a and b, and forms the quantities shown. Each butterfly requires one complex
multiplication and two complex additions.
By considering together the computations involving common output frequencies from
the two half-length DFTs, we see that the two complex multiplies are related to each
other, and we can reduce our computational work even further. By further
decomposing the length-4 DFTs into two length-2 DFTs and combining their outputs,
we arrive at the diagram summarizing the length-8 fast Fourier transform (Figure 5.12
224
(Length-8 DFT decomposition)). Although most of the complex multiplies are quite
simple (multiplying by
means swapping real and imaginary parts and changing their signs), let's count those
for purposes of evaluating the complexity as full complex multiplies. We have
complex multiplies and N =8 complex additions for each stage and log2N = 3 stages,
making the number of basic computations
as predicted.
Exercise 5.9.2
Note that the ordering of the input sequence in the two parts of Figure 5.12 (Length-8
DFT decomposition) aren't quite the same. Why not? How is the ordering determined?
Other "fast" algorithms were discovered, all of which make use of how many common
factors the transform length N has. In number theory, the number of prime factors a
given integer has measures how composite it is. The numbers 16 and 81 are highly
4
4
1
2
composite (equaling 2 and 3 respectively), the number 18 is less so (2 · 3 ), and 17
not at all (it's prime). In over thirty years of Fourier transform algorithm development,
the original Cooley-Tukey algorithm is far and away the most frequently used. It is so
computationally efficient that power-of-two transform lengths are frequently used
regardless of what the actual length of the data.
Exercise 5.9.3
Suppose the length of the signal were 500? How would you compute the spectrum of
this signal using the Cooley-Tukey algorithm? What would the length N of the
transform be?
5.11 Spectrograms
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We know how to acquire analog signals for digital processing (pre-filtering), sampling
(The Sampling Theorem (Page 201)), and A/D conversion (Analog-to-Digital Conversion
(Page 201)) and to compute spectra of discrete-time signals (using the FFT algorithm
(Fast Fourier Transform (FFT) (Page 221)), let's put these various components together
to learn how the spectrogram shown in Figure 5.14 (Speech Spectrogram), which is
used to analyze speech (Modeling the Speech Signal (Page 165)), is calculated. The
speech was sampled at a rate of 11.025 kHz and passed through a 16-bit A/D
converter.
POINT OF INTEREST: Music compact discs (CDs) encode their signals at a sampling rate
of 44.1 kHz. We'll learn the rationale for this number later. The 11.025 kHz sampling
rate for the speech is 1/4 of the CD sampling rate, and was the lowest available
225
sampling rate commensurate with speech signal bandwidths available on my
computer.
Exercise 5.10.1
Looking at Figure 5.14 (Speech Spectrogram) the signal lasted a little over 1.2 seconds.
How long was the sampled signal (in terms of samples)? What was the datarate during
the sampling process in bps (bits per second)? Assuming the computer storage is
organized in terms of bytes (8-bit quantities), how many bytes of computer memory
does the speech consume?
Figure 5.14 Speech Spectrogram
The resulting discrete-time signal, shown in the bottom of Figure 5.14 (Speech
Spectrogram), clearly changes its character with time. To display these spectral
changes, the long signal was sectioned into frames: comparatively short, contiguous
groups of samples. Conceptually, a Fourier transform of each frame is calculated using
the FFT. Each frame is not so long that signifcant signal variations are retained within a
frame, but not so short that we lose the signal's spectral character. Roughly speaking,
the speech signal's spectrum is evaluated over successive time segments and stacked
side by side so that the x-axis corresponds to time and the y-axis frequency, with color
indicating the spectral amplitude.
An important detail emerges when we examine each framed signal (Figure 5.15
(Spectrogram Hanning vs. Rectangular)).
226
Figure 5.15 Spectrogram Hanning vs. Rectangular
The top waveform is a segment 1024 samples long taken from the beginning of the
"Rice University" phrase. Computing Figure 5.14 (Speech Spectrogram) involved
creating frames, here demarked by the vertical lines, that were 256 samples long and
_nding the spectrum of each. If a rectangular window is applied (corresponding to
extracting a frame from the signal), oscillations appear in the spectrum (middle of
bottom row). Applying a Hanning window gracefully tapers the signal toward frame
edges, thereby yielding a more accurate computation of the signal's spectrum at that
moment of time.
At the frame's edges, the signal may change very abruptly, a feature not present in the
original signal. A transform of such a segment reveals a curious oscillation in the
spectrum, an artifact directly related to this sharp amplitude change. A better way to
frame signals for spectrograms is to apply a window: Shape the signal values within a
frame so that the signal decays gracefully as it nears the edges. This shaping is
accomplished by multiplying the framed signal by the sequence w (n). In sectioning
the signal, we essentially applied a rectangular window: w (n)=1, 0 ≤ n ≤ N −1. A much
more graceful window is the Hanning window; it has the cosine shape
As shown in Figure 5.15 (Spectrogram Hanning vs. Rectangular), this shaping greatly
reduces spurious oscillations in each frame's spectrum. Considering the spectrum of
the Hanning windowed frame, we find that the oscillations resulting from applying the
rectangular window obscured a formant (the one located at a little more than half the
Nyquist frequency).
Exercise 5.10.2
What might be the source of these oscillations? To gain some insight, what is the
length-2N discrete Fourier transform of a length-N pulse? The pulse emulates the
rectangular window, and certainly has edges. Compare your answer with the length-2N
transform of a length-N Hanning window.
227
Figure 5.16 Non-overlapping windows
In comparison with the original speech segment shown in the upper plot, the non-overlapped Hanning
windowed version shown below it is very ragged. Clearly, spectral information extracted from the bottom
plot could well miss important features present in the original.
If you examine the windowed signal sections in sequence to examine windowing's
effect on signal amplitude, we see that we have managed to amplitude-modulate the
signal with the periodically repeated window (Figure 5.16 (Non-overlapping windows)).
To alleviate this problem, frames are overlapped (typically by half a frame duration).
This solution requires more Fourier transform calculations than needed by
rectangular windowing, but the spectra are much better behaved and spectral
changes are much better captured.
The speech signal, such as shown in the speech spectrogram (Figure 5.14: Speech
Spectrogram), is sectioned into overlapping, equal-length frames, with a Hanning
window applied to each frame. The spectra of each of these is calculated, and
displayed in spectrograms with frequency extending vertically, window time location
running horizontally, and spectral magnitude color-coded. Figure 5.17 (Overlapping
windows for computing spectrograms) illustrates these computations.
Figure 5.17 Overlapping windows for computing spectrograms
The original speech segment and the sequence of overlapping Hanning windows
applied to it are shown in the upper portion. Frames were 256 samples long and a
228
Hanning window was applied with a half-frame overlap. A length-512 FFT of each
frame was computed, with the magnitude of the first 257 FFT values displayed
vertically, with spectral amplitude values color-coded.
Exercise 5.10.3
Why the specific values of 256 for N and 512 for K? Another issue is how was the
length-512 transform of each length-256 windowed frame computed?
5.12 Discrete-Time Systems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we developed analog systems, interconnecting the circuit elements provided a
natural starting place for constructing useful devices. In discrete-time signal
processing, we are not limited by hardware considerations but by what can be
constructed in software.
Exercise 5.11.1
One of the first analog systems we described was the amplifier (Amplifiers (Page 27):
Amplifiers). We found that implementing an amplifier was difficult in analog systems,
requiring an op-amp at least. What is the discrete-time implementation of an
amplifier? Is this especially hard or easy?
In fact, we will discover that frequency-domain implementation of systems, wherein
we multiply the input signal's Fourier transform by a frequency response, is not only a
viable alternative, but also a computationally efficient one. We begin with discussing
the underlying mathematical structure of linear, shift-invariant systems, and devise
how software filters can be constructed.
5.13 Discrete-Time Systems in the Time-Domain
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A discrete-time signal s(n) is delayed by n0 samples when we write s(n - n0), with n0 > 0.
Choosing n0 to be negative advances the signal along the integers. As opposed to
analog delays (Delay (Page 28) : Delay), discrete-time delays can only be integer
valued. In the frequency domain, delaying a signal corresponds to a linear phase shift
of the signal's discrete-time Fourier transform :
Linear discrete-time systems have the superposition property.
(5.40)
229
A discrete-time system is called shift-invariant (analogous to time-invariant analog
systems (Time-Invariant Systems (Page 30) ) if delaying the input delays the
corresponding output. If S (x (n)) = y (n), then a shift-invariant system has the property
(5.41)
We use the term shift-invariant to emphasize that delays can only have integer values
in discrete-time, while in analog signals, delays can be arbitrarily valued.
We want to concentrate on systems that are both linear and shift-invariant. It will be
these that allow us the full power of frequency-domain analysis and implementations.
Because we have no physical constraints in "constructing" such systems, we need only
a mathematical specification. In analog systems, the differential equation specifies the
input-output relationship in the time-domain. The corresponding discrete-time
specification is the difference equation.
(5.42)
Here, the output signal y (n) is related to its past values y (n - l), l = {1; … ; p}, and to the
current and past values of the input signal x (n). The system's characteristics are
determined by the choices for the number of coefficients p and q and the coefficients'
values {a1, ... , ap} and {b0, b1, ... , bq}.
ASIDE: There is an asymmetry in the coefficients: where is a0? This coefficient would
multiply the y(n) term in (5.42). We have essentially divided the equation by it, which does
not change the input-output relationship. We have thus created the convention that a0 is
always one.
As opposed to differential equations, which only provide an implicit description of a
system (we must somehow solve the differential equation), difference equations
provide an explicit way of computing the output for any input. We simply express the
difference equation by a program that calculates each output from the previous
output values, and the current and previous inputs.
Difference equations are usually expressed in software with for loops. A MATLAB
program that would compute the first 1000 values of the output has the form
for n=1:1000
y(n) = sum(a.*y(n-1:-1:n-p)) + sum(b.*x(n:-1:n-q));
end
An important detail emerges when we consider making this program work; in fact, as
written it has (at least) two bugs. What input and output values enter into the
computation of y (1)? We need values for y (0), y (−1), ..., values we have not yet
computed. To compute them, we would need more previous values of the output,
which we have not yet computed. To compute these values, we would need even
earlier values, ad infinitum. The way out of this predicament is to specify the system's
initial conditions: we must provide the p output values that occurred before the
230
input started. These values can be arbitrary, but the choice does impact how the
system responds to a given input. One choice gives rise to a linear system: Make
the initial conditions zero. The reason lies in the Definition of a linear system (Section
2.6.6: Linear Systems): The only way that the output to a sum of signals can be the sum of
the individual outputs occurs when the initial conditions in each case are zero.
Exercise 5.12.1
The initial condition issue resolves making sense of the difference equation for inputs
that start at some index. However, the program will not work because of a
programming, not conceptual, error. What is it? How can it be "fixed?"
Example 5.5
Let's consider the simple system having p =1 and q =0.
(5.43)
To compute the output at some index, this difference equation says we need to know
what the previous output y (n − 1) and what the input signal is at that moment of time.
In more detail, let's compute this system's output to a unit-sample input: x (n)= δ (n).
Because the input is zero for negative indices, we start by trying to compute the
output at n = 0.
(5.44)
What is the value of y (−1)? Because we have used an input that is zero for all negative
indices, it is reasonable to assume that the output is also zero. Certainly, the
difference equation would not describe a linear system (Section 2.6.6: Linear Systems)
if the input that is zero for all time did not produce a zero output. With this
assumption, y (−1) = 0, leaving y (0) = b. For n> 0, the input unit-sample is zero, which
leaves us with the difference equation y (n)= ay (n − 1) , n> 0 . We can envision how the
filter responds to this input by making a table.
(5.45)
231
n
x (n)
y (n)
−1
0
0
0
1
b
1
0
ba
2
0
ba2
:
0
:
n
0
ban
Coefficient values determine how the output behaves. The parameter b can be any
value, and serves as a gain. The effect of the parameter a is more complicated (Table).
If it equals zero, the output simply equals the input times the gain b. For all non-zero
values of a, the output lasts forever; such systems are said to be IIR (Infinite Impulse
Response). The reason for this terminology is that the unit sample also known as the
impulse (especially in analog situations), and the system's response to the "impulse"
lasts forever. If a is positive and less than one, the output is a decaying exponential.
When a =1, the output is a unit step. If a is negative and greater than −1, the output
oscillates while decaying exponentially. When a = −1, the output changes sign forever,
alternating between b and −b. More dramatic effects when |a| > 1; whether positive
or negative, the output signal becomes larger and larger, growing exponentially.
Figure 5.18
The input to the simple example system, a unit sample, is shown at the top, with the outputs for several
system parameter values shown below.
232
Positive values of a are used in population models to describe how population size
increases over time. Here, n might correspond to generation. The diference equation
says that the number in the next generation is some multiple of the previous one. If
this multiple is less than one, the population becomes extinct; if greater than one, the
population fourishes. The same diference equation also describes the effect of
compound interest on deposits. Here, n indexes the times at which compounding
occurs (daily, monthly, etc.), a equals the compound interest rate plus one, and b =1
(the bank provides no gain). In signal processing applications, we typically require that
the output remain bounded for any input. For our example, that means that we
restrict |a| < 1 and choose values for it and the gain according to the application.
Exercise 5.12.2
Note that the difference equation (5.42),
does not involve terms like y (n + 1) or x (n + 1) on the equation's right side. Can such
terms also be included? Why or why not?
Figure 5.19 The plot shows the unit-sample response of a length-5 boxcar filter.
Example 5.6
A somewhat different system has no "a" coefficients. Consider the difference equation
(5.46)
Because this system's output depends only on current and previous input values, we
need not be concerned with initial conditions. When the input is a unit-sample, the
output equals
for n = {0,...,q − 1}, then equals zero thereafter. Such systems are said to be FIR (Finite
Impulse Response) because their unit sample responses have finite duration. Plotting
this response (Figure 5.19) shows that the unit-sample response is a pulse of width q
and height
This waveform is also known as a boxcar, hence the name boxcar filter given to this
system. We'll derive its frequency response and develop its filtering interpretation in
the next section. For now, note that the difference equation says that each output
value equals the average of the input's current and previous values. Thus, the output
233
equals the running average of input's previous q values. Such a system could be used
to produce the average weekly temperature (q =7) that could be updated daily.
[MEDIA OBJECT] (This media object is a LabVIEW VI. Please view or download it at
<DiscreteTimeSys.llb>)
5.14 Discrete-Time Systems in the Frequency Domain
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
As with analog linear systems, we need to find the frequency response of discretetime systems. We used impedances to derive directly from the circuit's structure the
frequency response. The only structure we have so far for a discrete-time system is
the difference equation. We proceed as when we used impedances: let the input be a
complex exponential signal. When we have a linear, shift-invariant system, the output
should also be a complex exponential of the same frequency, changed in amplitude
and phase. These amplitude and phase changes comprise the frequency response we
j2πfn
seek. The complex exponential input signal is x (n)= Xe
. Note that this input
occurs for all values of n. No need to worry about initial conditions here. Assume the
output has a similar form: y (n)= Ye
difference equation (5.42), we have
j2πfn
. Plugging these signals into the fundamental
(5.47)
The assumed output does indeed satisfy the diference equation if the output complex
amplitude is related to the input amplitude by
This relationship corresponds to the system's frequency response or, by another
name, its transfer function. We find that any discrete-time system defined by a
diference equation has a transfer function given by
(5.48)
Furthermore, because any discrete-time signal can be expressed as a superposition of
complex exponential signals and because linear discrete-time systems obey the
Superposition Principle, the transfer function relates the discrete-time Fourier
transform of the system's output to the input's Fourier transform.
(5.49)
Example 5.7
234
The frequency response of the simple IIR system (difference equation given in a
previous example (Discrete-Time Systems in the Time-Domain (Page 228)) is given by
(5.50)
This Fourier transform occurred in a previous example; the exponential signal
spectrum (Figure 5.10: Spectra of exponential signals) portrays the magnitude and
phase of this transfer function. When the filter coefficient a is positive, we have a
lowpass filter; negative a results in a highpass filter. The larger the coefficient in
magnitude, the more pronounced the lowpass or highpass filtering.
Example 5.8
The length-q boxcar filter (difference equation found in a previous example (DiscreteTime Systems in the Time-Domain (Page 228)) has the frequency response
(5.51)
This expression amounts to the Fourier transform of the boxcar signal (5.13). There
we found that this frequency response has a magnitude equal to the absolute value of
dsinc(πf); see the length-10 filter's frequency response (Figure 5.11: Spectrum of
length-ten pulse). We see that boxcar filters length-q signal averages have a lowpass
behavior, having a cutoff frequency of
.
Exercise 5.13.1
Suppose we multiply the boxcar filter's coefficients by a sinusoid:
Use Fourier transform properties to determine the transfer function. How would you
characterize this system: Does it act like a filter? If so, what kind of filter and how do
you control its characteristics with the filter's coefficients?
These examples illustrate the point that systems described (and implemented) by
difference equations serve as filters for discrete-time signals. The filter's order is given
by the number p of denominator coefficients in the transfer function (if the system is
IIR) or by the number q of numerator coefficients if the filter is FIR. When a system's
transfer function has both terms, the system is usually IIR, and its order equals
p regardless of q. By selecting the coefficients and filter type, filters having virtually
any frequency response desired can be designed. This design flexibility can't be found
in analog systems. In the next section, we detail how analog signals can be filtered by
computers, offering a much greater range of filtering possibilities than is possible with
circuits.
235
5.14.1 Filtering in the Frequency Domain
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Because we are interested in actual computations rather than analytic calculations, we
must consider the details of the discrete Fourier transform. To compute the
length-N DFT, we assume that the signal has a duration less than or equal to N.
Because frequency responses have an explicit frequency-domain specification (5.47) in
terms of filter coefficients, we don't have a direct handle on which signal has a Fourier
transform equaling a given frequency response. Finding this signal is quite easy. First
of all, note that the discrete-time Fourier transform of a unit sample equals one for all
frequencies. Since the input and output of linear, shift-invariant systems are related to
each other by
a unit-sample input, which has X (e
j2πf
) = 1, results in the output's Fourier
transform equaling the system's transfer function.
Exercise 5.14.1
This statement is a very important result. Derive it yourself. In the time-domain, the
output for a unit-sample input is known as the system's unit-sample response, and is
denoted by h (n). Combining the frequency-domain and time-domain interpretations
of a linear, shift-invariant system's unit-sample response, we have that h (n) and the
transfer function are Fourier transform pairs in terms of the discrete-time Fourier
transform.
(5.52)
Returning to the issue of how to use the DFT to perform filtering, we can analytically
specify the frequency response, and derive the corresponding length-N DFT by
sampling the frequency response.
(5.53)
Computing the inverse DFT yields a length-N signal no matter what the actual
duration of the unit-sample response might be. If the unit-sample response has a
duration less than or equal to N (it's a FIR filter), computing the inverse DFT of the
sampled frequency response indeed yields the unit-sample response. If, however, the
duration exceeds N, errors are encountered. The nature of these errors is easily
explained by appealing to the Sampling Theorem. By sampling in the frequency
domain, we have the potential for aliasing in the time domain (sampling in one
domain, be it time or frequency, can result in aliasing in the other) unless we sample
fast enough. Here, the duration of the unit-sample response determines the minimal
sampling rate that prevents aliasing. For FIR systems they by Definition have finiteduration unit sample responses the number of required DFT samples equals the unitsample response's duration: N ≥ q.
236
Exercise 5.14.2
Derive the minimal DFT length for a length-q unit-sample response using the Sampling
Theorem. Because sampling in the frequency domain causes repetitions of the unitsample response in the time domain, sketch the time-domain result for various
choices of the DFT length N.
Exercise 5.14.3
Express the unit-sample response of a FIR filter in terms of difference equation
coefficients. Note that the corresponding question for IIR filters is far more difficult to
answer: Consider the example (Discrete-Time Systems in the Time-Domain (Page
228)).
For IIR systems, we cannot use the DFT to find the system's unit-sample response:
aliasing of the unit-sample response will always occur. Consequently, we can only
implement an IIR filter accurately in the time domain with the system's difference
equation. Frequency-domain implementations are restricted to FIR filters.
Another issue arises in frequency-domain filtering that is related to time-domain
aliasing, this time when we consider the output. Assume we have an input signal
having duration Nx that we pass through a FIR filter having a length-q +1 unit-sample
response. What is the duration of the output signal? The difference equation for this
filter is
(5.54)
This equation says that the output depends on current and past input values, with the
input value q samples previous defining the extent of the filter's memory of past
input values. For example, the output at index Nx depends on x (Nx) (which equals
zero), x (Nx − 1), through x (Nx − q). Thus, the output returns to zero only after the last
input value passes through the filter's memory. As the input signal's last value occurs
at index Nx − 1, the last nonzero output value occurs when n − q = Nx − 1 or n = q + Nx
− 1. Thus, the output signal's duration equals q + Nx.
Exercise 5.14.4
In words, we express this result as "The output's duration equals the input's duration
plus the filter's duration minus one.". Demonstrate the accuracy of this statement.
The main theme of this result is that a filter's output extends longer than either its
input or its unit-sample response. Thus, to avoid aliasing when we use DFTs, the
dominant factor is not the duration of input or of the unit-sample response, but of the
output. Thus, the number of values at which we must evaluate the frequency
response's DFT must be at least q + Nx and we must compute the same length DFT of
the input. To accommodate a shorter signal than DFT length, we simply zero-pad the
input: Ensure that for indices extending beyond the signal's duration that the signal is
zero. Frequency-domain filtering, diagrammed in Figure 5.20, is accomplished by
storing the filter's frequency response as the DFT H (k), computing the input's DFT X
(k), multiplying them to create the output's DFT Y (k) = H (k) X (k), and computing the
inverse DFT of the result to yield y (n).
237
Figure 5.20
To filter a signal in the frequency domain, first compute the DFT of the input, multiply the result by the
sampled frequency response, and fnally compute the inverse DFT of the product. The DFT's length must
be at least the sum of the input's and unit-sample response's duration minus one. We calculate these
discrete Fourier transforms using the fast Fourier transform algorithm, of course.
Before detailing this procedure, let's clarify why so many new issues arose in trying to
develop a frequency-domain implementation of linear filtering. The frequency-domain
relationship between a filter's input and output is always true:
The Fourier transforms in this result are discrete time Fourier transforms; for
example,
Unfortunately, using this relationship to perform filtering is restricted to the situation
when we have analytic formulas for the frequency response and the input signal. The
reason why we had to "invent" the discrete Fourier transform (DFT) has the same
origin: The spectrum resulting from the discrete-time Fourier transform depends on
the continuous frequency variable f. That's fine for analytic calculation, but
computationally we would have to make an uncountably infinite number of
computations.
Note: Did you know that two kinds of infinities can be meaningfully defined? A
countablyinfinite quantity means that it can be associated with a limiting
process associated with integers. An uncountablyinfinite quantity cannot be
so associated. The number of rational numbers is countably infinite (the
numerator and denominator correspond to locating the rational by row and
column; the total number so-located can be counted, voila!); the number of
irrational numbers is uncountably infinite. Guess which is "bigger?"
The DFT computes the Fourier transform at a finite set of frequencies samples the
true spectrum which can lead to aliasing in the time-domain unless we sample
sufficiently fast. The sampling interval here is
for a length-K DFT: faster sampling to avoid aliasing thus requires a longer transform
calculation. Since the longest signal among the input, unit-sample response and
output is the output, it is that signal's duration that determines the transform length.
We simply extend the other two signals with zeros (zero-pad) to compute their DFTs.
Example 5.9
238
Suppose we want to average daily stock prices taken over last year to yield a running
weekly average (average over five trading sessions). The filter we want is a length-5
averager (as shown in the unit-sample response (Figure 5.19), and the input's duration
is 253 (365 calendar days minus weekend days and holidays). The output duration will
be 253 + 5 - 1 = 257, and this determines the transform length we need to use.
Because we want to use the FFT, we are restricted to power-of-two transform lengths.
We need to choose any FFT length that exceeds the required DFT length. As it turns
8
out, 256 is a power of two (2 = 256), and this length just undershoots our required
length. To use frequency domain techniques, we must use length-512 fast Fourier
transforms.
Figure 5.21 .
The blue line shows the Dow Jones Industrial Average from 1997, and the red one the length-5 boxcarfiltered result that provides a running weekly of this market index. Note the "edge" effects in the filtered
output
Figure 5.21 shows the input and the filtered output. The MATLAB programs that
compute the filtered output in the time and frequency domains are
Time Domain
h = [1 1 1 1 1]/5;
y = filter(h,1,[djia zeros(1,4)]);
Frequency Domain
h = [1 1 1 1 1]/5;
DJIA = fft(djia, 512);
H = fft(h, 512);
Y = H.*X;
y = ifft(Y);
Note: The filter program has the feature that the length of its output equals the
length of its input. To force it to produce a signal having the proper length, the
program zero-pads the input appropriately.
239
MATLAB's
fft
function automatically zero-pads its input if the specified transform
length (its second argument) exceeds the signal's length. The frequency domain result
−11
will have a small imaginary component largest value is 2.2 × 10
because of the
inherent finite precision nature of computer arithmetic. Because of the unfortunate
misfit between signal lengths and favored FFT lengths, the number of arithmetic
operations in the time-domain implementation is far less than those required by the
frequency domain version: 514 versus 62,271. If the input signal had been one sample
shorter, the frequency-domain computations would have been more than a factor of
two less (28,696), but far more than in the time-domain implementation.
An interesting signal processing aspect of this example is demonstrated at the
beginning and end of the output. The ramping up and down that occurs can be traced
to assuming the input is zero before it begins and after it ends. The filter "sees" these
initial and final values as the difference equation passes over the input. These artifacts
can be handled in two ways: we can just ignore the edge effects or the data from
previous and succeeding years' last and first week, respectively, can be placed at the
ends.
5.15 Efficiency of Frequency-Domain Filtering
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
To determine for what signal and filter durations a time-or frequency-domain
implementation would be the most efficient, we need only count the computations
required by each. For the time-domain, difference equation approach, we need Nx (2
(q) + 1). The frequency-domain approach requires three Fourier transforms, each
requiring
computations for a length-K FFT, and the multiplication of two spectra
(6K computations). The output-signal-duration-determined length must be at least
Nx + q. Thus, we must compare
Exact analytic evaluation of this comparison is quite difficult (we have a transcendental
equation to solve). Insight into this comparison is best obtained by dividing by Nx.
With this manipulation, we are evaluating the number of computations per sample.
For any given value of the filter's order q, the right side, the number of frequencydomain computations, will exceed the left if the signal's duration is long enough.
However, for filter durations greater than about 10, as long as the input is at least 10
samples, the frequency-domain approach is faster so long as the FFT's power-of-two
constraint is advantageous.
240
The frequency-domain approach is not yet viable; what will we do when the input
signal is infinitely long? The difference equation scenario fits perfectly with the
envisioned digital filtering structure (Figure 5.24), but so far we have required the
input to have limited duration (so that we could calculate its Fourier transform). The
solution to this problem is quite simple: Section the input into frames, filter each, and
add the results together. To section a signal means expressing it as a linear
combination of length-Nx non-overlapping "chunks." Because the filter is linear,
filtering a sum of terms is equivalent to summing the results of filtering each term.
(5.55)
As illustrated in Figure 5.22, note that each filtered section has a duration longer than
the input. Conse quently, we must literally add the filtered sections together, not just
butt them together.
Figure 5.22
The noisy input signal is sectioned into length-48 frames, each of which is filtered using frequency-domain
techniques. Each filtered section is added to other outputs that overlap to create the signal equivalent to
having filtered the entire input. The sinusoidal component of the signal is shown as the red dashed line.
Computational considerations reveal a substantial advantage for a frequency-domain
implementation over a time-domain one. The number of computations for a timedomain implementation essentially remains constant whether we section the input or
not. Thus, the number of computations for each output is 2(q)+1. In the frequencydomain approach, computation counting changes because we need only compute the
filter's frequency response H (k) once, which amounts to a fixed overhead. We need
241
only compute two DFTs and multiply them to filter a section. Letting Nx denote a
section's length, the number of computations for a section amounts to
In addition, we must add the filtered outputs together; the number of terms to add
corresponds to the excess duration of the output compared with the input (q). The
frequency-domain approach thus requires
computations per output value. For even modest filter orders, the frequency-domain
approach is much faster.
Exercise 5.15.1
Show that as the section length increases, the frequency domain approach becomes
increasingly more efficient. Note that the choice of section duration is arbitrary. Once
the filter is chosen, we should section so that the required FFT length is precisely a
L
power of two: Choose Nx so that Nx + q = 2 .
Implementing the digital filter shown in the A/D block diagram (Figure 5.24) with a
frequency-domain implementation requires some additional signal management not
required by time-domain implementations. Conceptually, a real-time, time-domain
filter could accept each sample as it becomes available, calculate the difference
equation, and produce the output value, all in less than the sampling interval Ts.
Frequency-domain approaches don't operate on a sample-by-sample basis; instead,
they operate on sections. They filter in real time by producing Nx outputs for the same
number of inputs faster than NxTs. Because they generally take longer to produce an
output section than the sampling interval duration, we must filter one section while
accepting into memory the next section to be filtered. In programming, the operation
of building up sections while computing on previous ones is known as buffering.
Buffering can also be used in time-domain filters as well but isn't required.
Example 5.10
We want to lowpass filter a signal that contains a sinusoid and a significant amount of
noise. The example shown in Figure 5.22 shows a portion of the noisy signal's
waveform. If it weren't for the overlaid sinusoid, discerning the sine wave in the signal
is virtually impossible. One of the primary applications of linear filters is noise
removal: preserve the signal by matching filter's passband with the signal's spectrum
and greatly reduce all other frequency components that may be present in the noisy
signal.
A smart Rice engineer has selected a FIR filter having a unit-sample response
corresponding a
period-17 sinusoid:
which makes q = 16. Its frequency response (determined by computing the discrete
Fourier transform) is shown in Figure 5.23. To apply, we can select the length of each
section so that the frequency-domain filtering approach is maximally efficient: Choose
242
the section length Nx so that Nx+ q is a power of two. To use a length-64 FFT, each
section must be 48 samples long. Filtering with the difference equation would require
33 computations per output while the frequency domain requires a little over 16; this
frequency-domain implementation is over twice as fast! Figure 5.22 shows how
frequency-domain filtering works.
Figure 5.23
The figure shows the unit-sample response of a length-17 Hanning filter on the left and the frequency
response on the right. This filter functions as a lowpass filter having a cutoff frequency of about 0.1
We note that the noise has been dramatically reduced, with a sinusoid now clearly
visible in the filtered output. Some residual noise remains because noise components
within the filter's passband appear in the output as well as the signal.
Exercise 5.15.2
Note that when compared to the input signal's sinusoidal component, the output's
sinusoidal component seems to be delayed. What is the source of this delay? Can it be
removed?
5.16 Discrete-Time Filtering of Analog Signals
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Because of the Sampling Theorem (The Sampling Theorem (Page 202)), we can
process, in particular filter, analog signals "with a computer" by constructing the
system shown in Figure 5.24. To use this system, we are assuming that the input signal
has a lowpass spectrum and can be bandlimited without afecting important signal
aspects. Bandpass signals can also be filtered digitally, but require a more complicated
system. Highpass signals cannot be filtered digitally. Note that the input and output
filters must be analog filters; trying to operate without them can lead to potentially
very inaccurate digitization.
243
Figure 5.24 A system
To process an analog signal digitally, the signal x (t) must be filtered with an anti-aliasing filter (to ensure a
bandlimited signal) before A/D conversion. This lowpass filter (LPF) has a cutoff frequency of W Hz, which
determines allowable sampling intervals Ts. The greater the number of bits in the amplitude quantization
portion Q [·] of the A/D converter, the greater the accuracy of the entire system. The resulting digital
signal x (n) can now be filtered in the time-domain with a difference equation or in the frequency domain
with Fourier transforms. The resulting output y (n) then drives a D/A converter and a second anti-aliasing
filter (having the same bandwidth as the first one).
Another implicit assumption is that the digital filter can operate in real time: The
computer and the filtering algorithm must be sufficiently fast so that outputs are
computed faster than input values arrive. The sampling interval, which is determined
by the analog signal's bandwidth, thus determines how long our program has to
compute each output y (n). The computational complexity for calculating each output
with a difference equation (5.42) is O (p + q). Frequency domain implementation of the
filter is also possible. The idea begins by computing the Fourier transform of a lengthN portion of the input x (n), multiplying it by the filter's transfer function, and
computing the inverse transform of the result. This approach seems overly complex
and potentially inefficient. Detailing the complexity, however, we have O(NlogN) for the
two transforms (computed using the FFT algorithm) and O (N) for the multiplication by
the transfer function, which makes the total complexity O(NlogN) for N input values. A
frequency domain implementation thus requires O(NlogN) computational complexity
for each output value. The complexities of time-domain and frequency-domain
implementations depend on different aspects of the filtering: The time-domain
implementation depends on the combined orders of the filter while the frequencydomain implementation depends on the logarithm of the Fourier transform's length.
It could well be that in some problems the time-domain version is more efficient
(more easily satisfies the real time requirement), while in others the frequency domain
approach is faster. In the latter situations, it is the FFT algorithm for computing the
Fourier transforms that enables the superiority of frequency-domain
implementations. Because complexity considerations only express how algorithm
running-time increases with system parameter choices, we need to detail both
implementations to determine which will be more suitable for any given filtering
problem. Filtering with a difference equation is straightforward, and the number of
computations that must be made for each output value is 2 (p+q).
Exercise 5.16.1
Derive this value for the number of computations for the general difference equation
(5.42).
244
5.17 Digital Signal Processing Problems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Problem 5.1: Sampling and Filtering
The signal s (t) is bandlimited to 4 kHz. We want to sample it, but it has been subjected
to various signal processing manipulations.
1. What sampling frequency (if any works) can be used to sample the result of
passing s (t) through an RC highpass filter with R = 10kΩ and C = 8nF ?
2. What sampling frequency (if any works) can be used to sample the derivative of s
(t)?
3. The signal s (t) has been modulated by an 8 kHz sinusoid having an unknown
phase: the resulting signal is s (t) sin (2πf0t + Ï•) , with f0 = 8kHz and Ï• =? Can the
modulated signal be sampled so that the original signal can be recovered from
the modulated signal regardless of the phase value φ? If so, show how and find
the smallest sampling rate that can be used; if not, show why not.
Problem 5.2: Non-Standard Sampling
Using the properties of the Fourier series can ease finding a signal's spectrum.
1. Suppose a signal s (t) is periodic with period T . If ck represents the signal's Fourier
series coefficients, what are the Fourier series coefficients of
2. Find the Fourier series of the signal p (t) shown in Figure 5.25 (Pulse Signal).
3. Suppose this signal is used to sample a signal bandlimited to
Find an expression for and sketch the spectrum of the sampled signal.
4. Does aliasing occur? If so, can a change in sampling rate prevent aliasing; if not,
show how the signal can be recovered from these samples.
Figure 5.25 Pulse Signal
Problem 5.3: A Different Sampling Scheme
245
A signal processing engineer from Texas A&M claims to have developed an improved
sampling scheme. He multiplies the bandlimited signal by the depicted periodic pulse
signal to perform sampling (Figure 5.26).
Figure 5.26
1. Find the Fourier spectrum of this signal.
2. Will this scheme work? If so, how should TS be related to the signal's bandwidth? If
not, why not?
Problem 5.4: Bandpass Sampling
The signal s(t) has the indicated spectrum.
Figure 5.27
1. What is the minimum sampling rate for this signal suggested by the Sampling
Theorem?
2. Because of the particular structure of this spectrum, one wonders whether a
lower sampling rate could be used. Show that this is indeed the case, and find the
system that reconstructs s (t) from its samples.
Problem 5.5: Sampling Signals
If a signal is bandlimited to W Hz, we can sample it at any rate
and recover the waveform exactly. This statement of the Sampling Theorem can be
taken to mean that all information about the original signal can be extracted from the
samples. While true in principle, you do have to be careful how you do so. In addition
to the rms value of a signal, an important aspect of a signal is its peak value, which
equals max {|s (t) |} .
246
1.
Let s (t) be a sinusoid having frequency W Hz. If we sample it at
precisely the Nyquist rate, how accurately do the samples
convey the sinusoid's amplitude? In other words, find the
worst case example.
2.
How fast would you need to sample for the amplitude estimate
to be within 5% of the true value?
Another issue in sampling is the inherent amplitude
quantization produced by A/D converters. Assume the
maximum voltage allowed by the converter is Vmax volts and
that it quantizes amplitudes to b bits. We can express the
quantized sample
3.
as
where
represents the quantization error at the nth sample. Assuming
the converter rounds, how large is maximum quantization
error?
4.
We can describe the quantization error as noise, with a power
proportional to the square of the maximum error. What is the
signal-to-noise ratio of the quantization error for a full-range
sinusoid? Express your result in decibels.
Problem 5.6: Hardware Error
An A/D converter has a curious hardware problem: Every other sampling pulse is half
its normal amplitude (Figure 5.28).
Figure 5.28
1. Find the Fourier series for this signal.
2. Can this signal be used to sample a bandlimited signal having highest frequency
Problem 5.7: Simple D/A Converter
Commercial digital-to-analog converters don't work this way, but a simple circuit
illustrates how they work. Let's assume we have a B-bit converter. Thus, we want to
convert numbers having a B-bit representation into a voltage proportional to that
247
number. The first step taken by our simple converter is to represent the number by a
sequence of B pulses occurring at multiples of a time interval T . The presence of a
pulse indicates a "1" in the corresponding bit position, and pulse absence means a "0"
occurred. For a 4-bit converter, the number 13 has the binary representation 1101
3
2
1
0
(1310 =1 × 2 +1 × 2 +0 × 2 +1 × 2 ) and would be represented by the depicted pulse
sequence. Note that the pulse sequence is "backwards" from the binary
representation. We'll see why that is.
Figure 5.29
This signal (Figure 5.29) serves as the input to a first-order RC lowpass filter. We want
to design the filter and the parameters Δ and T so that the output voltage at time
4T (for a 4-bit converter) is proportional to the number. This combination of pulse
creation and filtering constitutes our simple D/A converter. The requirements are
• The voltage at time t=4T should diminish by a factor of 2 the further the pulse
occurs from this time. In other words, the voltage due to a pulse at 3T should be
twice that of a pulse produced at 2T , which in turn is twice that of a pulse at T ,
etc.
• The 4-bit D/A converter must support a 10 kHz sampling rate.
Show the circuit that works. How do the converter's parameters change with sampling
rate and number of bits in the converter?
Problem 5.8: Discrete-Time Fourier Transforms Find the Fourier transforms of the
following sequences, where s (n) is some sequence having Fourier transform
a.
b.
c.
d.
Problem 5.9: Spectra of Finite-Duration Signals
Find the indicated spectra for the following signals.
248
1. The discrete-time Fourier transform of
2. The discrete-time Fourier transform of
3. The discrete-time Fourier transform of
4. The length-8 DFT of the previous signal.
Problem 5.10: Just Whistlin’
Sammy loves to whistle and decides to record and analyze his whistling in lab. He is a
very good whistler; his whistle is a pure sinusoid that can be described by sa (t) = sin
(4000t) . To analyze the spectrum, he samples his recorded whistle with a sampling
interval of TS =2.5 × 10
−4
to obtain s (n)= sa (nTS) . Sammy (wisely) decides to analyze a
few samples at a time, so he grabs 30 consecutive, but arbitrarily chosen, samples. He
calls this sequence x (n) and realizes he can write it as
x (n) = sin (4000nTS + θ) , n = {0,..., 29}
1. Did Sammy under-or over-sample his whistle?
2. What is the discrete-time Fourier transform of x (n) and how does it depend on θ?
3. How does the 32-point DFT of x (n) depend on θ?
Problem 5.11: Discrete-Time Filtering
We can find the input-output relation for a discrete-time filter much more easily than
for analog filters. The key idea is that a sequence can be written as a weighted linear
combination of unit samples.
1. Show that
w here δ (n) is the unit-sample.
1. If h (n) denotes the unit-sample response the output of a discrete-time linear,
shift-invariant filter to a unit-sample input find an expression for the output.
2. In particular, assume our filter is FIR, with the unit-sample response having
duration q +1. If the input has duration N, what is the duration of the filter's
output to this signal?
3. Let the filter be a boxcar averager:
for n={0,...,q} and zero otherwise. Let the input be a pulse of unit height and
duration N. Find the filter's output when
249
q an odd integer.
Problem 5.12: A Digital Filter
A digital filter has the depicted (Figure 5.30) unit-sample response.
Figure 5.30 A Digital Filter
1. What is the difference equation that defines this filter's input-output relationship?
2. What is this filter's transfer function?
3. What is the filter's output when the input is
Problem 5.13: A Special Discrete-Time Filter
Consider a FIR filter governed by the difference equation
1. Find this filter's unit-sample response.
2. Find this filter's transfer function. Characterize this transfer function (i.e., what
classic filter category does it fall into).
3. Suppose we take a sequence and stretch it out by a factor of three.
Sketch the sequence x (n) for some example s (n). What is the filter's output to this
input? In particular, what is the output at the indices where the input x (n) is
intentionally zero? Now how would you characterize this system?
Problem 5.14: Simulating the Real World
Much of physics is governed by differential equations, and we want to use signal
processing methods to simulate physical problems. The idea is to replace the
derivative with a discrete-time approximation and solve the resulting differential
equation. For example, suppose we have the differential equation
and we approximate the derivative by
250
where T essentially mounts to a sampling interval.
1. What is the diference equation that must be solved to approximate the
differential equation?
2. When x (t)= u (t) , the unit step, what will be the simulated output?
3. Assuming x (t) is a sinusoid, how should the sampling interval T be chosen so that
the approximation works well?
Problem 5.15: Derivatives
The derivative of a sequence makes little sense, but still, we can approximate it. The
digital filter described by the difference equation
y (n)= x (n) − x (n − 1)
resembles the derivative formula. We want to explore how well it works.
1. What is this filter's transfer function?
2. What is the filter's output to the depicted triangle input (Figure 5.31)?
Figure 5.31
1. Suppose the signal x (n) is a sampled analog signal: x (n) = x(nTs). Under what
conditions will the filter act like a differentiator? In other words, when will y (n) be
proportional to
Problem 5.16: The DFT
Let's explore the DFT and its properties.
1. What is the length-K DFT of length-N boxcar sequence, where N < K?
2. Consider the special case where K = 4. Find the inverse DFT of the product of the
DFTs of two length-3 boxcars.
3. If we could use DFTs to perform linear filtering, it should be true that the product
of the input's DFT and the unit-sample response's DFT equals the output's DFT.
So that you can use what you just calculated, let the input be a boxcar signal and
the unit-sample response also be a boxcar. The result of part (b) would then be
the filter's output if we could implement the filter with length-4 DFTs. Does the
actual output of the boxcar-filter equal the result found in the previous part?
4. What would you need to change so that the product of the DFTs of the input and
unit-sample response in this case equaled the DFT of the filtered output?
Problem 5.17: DSP Tricks
251
Sammy is faced with computing lots of discrete Fourier transforms. He will, of course,
use the FFT algorithm, but he is behind schedule and needs to get his results as
quickly as possible. He gets the idea of computing two transforms at one time by
computing the transform of s (n)= s1 (n)+ js2 (n) , where s1 (n) and s2 (n) are two realvalued signals of which he needs to compute the spectra. The issue is whether he can
retrieve the individual DFTs from the result or not.
1. What will be the DFT S (k) of this complex-valued signal in terms of S1 (k) and
S2 (k), the DFTs of the original signals?
2. Sammy's friend, an Aggie who knows some signal processing, says that retrieving
the wanted DFTs is easy: "Just find the real and imaginary parts of S (k)." Show
that this approach is too simplistic.
3. While his friend's idea is not correct, it does give him an idea. What approach will
work? Hint: Use the symmetry properties of the DFT.
4. How does the number of computations change with this approach? Will Sammy's
idea ultimately lead to a faster computation of the required DFTs?
Problem 5.18: Discrete Cosine Transform (DCT)
The discrete cosine transform of a length-N sequence is defined to be
Note that the number of frequency terms is 2N − 1: k = {0,..., 2N − 1}.
1. Find the inverse DCT.
2. Does a Parseval's Theorem hold for the DCT?
3. You choose to transmit information about the signal s (n) according to the DCT
coefficients. You could only send one, which one would you send?
Problem 5.19: A Digital Filter
A digital filter is described by the following difference equation:
1. What is this filter's unit sample response?
2. What is this filter's transfer function?
3. What is this filter's output when the input is
Problem 5.20: Another Digital Filter
A digital filter is determined by the following difference equation.
y (n)= y (n − 1) + x (n) − x (n − 4)
a) Find this filter's unit sample response.
b) What is the filter's transfer function? How would you characterize this filter
(lowpass, highpass, special purpose, ...)?
c) Find the filter's output when the input is the sinusoid
252
d) In another case, the input sequence is zero for n < 0, then becomes nonzero.
Sammy measures the output to be y (n)= δ (n)+ δ (n − 1) . Can his measurement be
correct? In other words, is there an input that can yield this output? If so, find the input
x (n) that gives rise to this output. If not, why not?
Problem 5.21: Yet Another Digital Filter
A filter has an input-output relationship given by the difference equation
1. What is the filter's transfer function? How would you characterize it?
2. What is the filter's output when the input equals
3. What is the filter's output when the input is the depicted discrete-time square
wave (Figure 5.32)?
Figure 5.32
Problem 5.22: A Digital Filter in the Frequency Domain
We have a filter with the transfer function
operating on the input signal x (n)= δ (n) − δ (n − 2) that yields the output y (n).
a) What is the filter's unit-sample response?
b) What is the discrete-Fourier transform of the output?
c) What is the time-domain expression for the output?
Problem 5.23: Digital Filters
A discrete-time system is governed by the difference equation
1. Find the transfer function for this system.
2. What is this system's output when the input is
253
3. If the output is observed to be y (n)= δ (n)+ δ (n − 1), then what is the input?
Problem 5.24: Digital Filtering
A digital filter has an input-output relationship expressed by the diference equation
1. Plot the magnitude and phase of this filter's transfer function.
2. What is this filter's output when
Problem 5.25: Detective Work
The signal x (n) equals δ (n) − δ (n − 1).
1. Find the length-8 DFT (discrete Fourier transform) of this signal.
2. You are told that when x (n) served as the input to a linear FIR (finite impulse
response) filter, the output was y (n)= δ (n) −δ (n − 1) + 2δ (n − 2) . Is this statement
true? If so, indicate why and find the system's unit sample response; if not, show
why not.
Problem 5.26:
A discrete-time, shift invariant, linear system produces an output y (n)= {1, −1, 0, 0,... }
when its input x (n) equals a unit sample.
1. Find the difference equation governing the system.
2. Find the output when x (n) = cos (2πf0n).
3. How would you describe this system's function?
Problem 5.27: Time Reversal has Uses
A discrete-time system has transfer function H(e
j2πf
) . A signal x (n) is passed through
this system to yield the signal w (n). The time-reversed signal w (−n) is then passed
through the system to yield the time-reversed output y (-n). What is the transfer
function between x (n) and y (n)?
Problem 5.28: Removing "Hum"
The slang word "hum" represents power line waveforms that creep into signals
because of poor circuit construction. Usually, the 60 Hz signal (and its harmonics) are
added to the desired signal. What we seek are filters that can remove hum. In this
problem, the signal and the accompanying hum have been sampled; we want to
design a digital filter for hum removal.
1. Find filter coefficients for the length-3 FIR filter that can remove a sinusoid having
digital frequency f0 from its input.
2. Assuming the sampling rate is fs to what analog frequency does f0 correspond?
3. A more general approach is to design a filter having a frequency response
magnitude proportional to
254
4. the absolute value of a cosine: |H(e
j2πf
) |∝|cos (πfN) | . In this way, not only can
the fundamental but also its first few harmonics be removed. Select the
parameter N and the sampling rate so that the frequencies at which the cosine
equals zero correspond to 60 Hz and its odd harmonics through the fifth.
5. Find the difference equation that defines this filter.
Problem 5.29: Digital AM Receiver
Thinking that digital implementations are always better, our clever engineer wants to
design a digital AM receiver. The receiver would bandpass the received signal, pass the
result through an A/D converter, perform all the demodulation with digital signal
processing systems, and end with a D/A converter to produce the analog message
signal. Assume in this problem that the carrier frequency is always a large even
multiple of the message signal's bandwidth W.
1. What is the smallest sampling rate that would be needed?
2. Show the block diagram of the least complex digital AM receiver.
3. Assuming the channel adds white noise and that a b-bit A/D converter is used,
what is the output's signal-to-noise ratio?
Problem 5.30: DFTs
A problem on Samantha's homework asks for the 8-point DFT of the discrete-time
signal δ (n − 1) + δ (n − 7).
1. What answer should Samantha obtain?
2. As a check, her group partner Sammy says that he computed the inverse DFT of
her answer and got δ (n + 1) + δ (n − 1). Does Sammy's result mean that
Samantha's answer is wrong?
3. The homework problem says to lowpass-filter the sequence by multiplying its DFT
by
and then computing the inverse DFT. Will this filtering algorithm work? If so, find the
filtered output; if not, why not?
Problem 5.31: Stock Market Data Processing
Because a trading week lasts five days, stock markets frequently compute running
averages each day over the previous five trading days to smooth price fluctuations.
The technical stock analyst at the Buy-Lo Sell-Hi brokerage firm has heard that FFT
filtering techniques work better than any others (in terms of producing more accurate
averages).
1. What is the diference equation governing the fve-day averager for daily stock
prices?
2. Design an efcient FFT-based filtering algorithm for the broker. How much data
should be processed at once to produce an efcient algorithm? What length
transform should be used?
255
3. Is the analyst's information correct that FFT techniques produce more accurate
averages than any others? Why or why not?
Problem 5.32: Echoes
Echoes not only occur in canyons, but also in auditoriums and telephone circuits. In
one situation where the echoed signal has been sampled, the input signal x
(n) emerges as x (n)+ a1x (n − n1)+ a2x (n − n2).
1. Find the difference equation of the system that models the production of echoes.
2. To simulate this echo system, ELEC 241 students are asked to write the most
efficient (quickest) program that has the same input-output relationship. Suppose
the duration of x(n) is 1,000 and that
Half the class votes to just program the difference equation while the other half
votes to program a frequency domain approach that exploits the speed of the
FFT. Because of the undecided vote, you must break the tie. Which approach is
more efficient and why?
3. Find the transfer function and difference equation of the system that suppresses
the echoes. In other words, with the echoed signal as the input, what system's
output is the signal x(n)?
Problem 5.33: Digital Filtering of Analog Signals RU Electronics wants to develop a filter
that would be used in analog applications, but that is implemented digitally. The filter
is to operate on signals that have a 10 kHz bandwidth, and will serve as a lowpass
filter.
1. What is the block diagram for your filter implementation? Explicitly denote which
components are analog, which are digital (a computer performs the task), and
which interface between analog and digital worlds.
2. What sampling rate must be used and how many bits must be used in the A/D
converter for the acquired signal's signal-to-noise ratio to be at least 60 dB? For
this calculation, assume the signal is a sinusoid.
3. If the filter is a length-128 FIR filter (the duration of the filter's unit-sample
response equals 128), should it be implemented in the time or frequency
domain?
4. Assuming H(e
j2πf
) is the transfer function of the digital filter, what is the transfer
function of your system?
Problem 5.34: Signal Compression
Because of the slowness of the Internet, lossy signal compression becomes important
if you want signals to be received quickly. An enterprising 241 student has proposed a
scheme based on frequency-domain processing. First of all, he would section the
signal into length-N blocks, and compute its N-point DFT. He then would discard (zero
the spectrum) at half of the frequencies, quantize them to b-bits, and send these over
the network. The receiver would assemble the transmitted spectrum and compute the
inverse DFT, thus reconstituting an N-point block.
256
a). At what frequencies should the spectrum be zeroed to minimize the error in this
lossy compression scheme?
b).The nominal way to represent a signal digitally is to use simple b-bit quantization of
the time-domain waveform. How long should a section be in the proposed scheme so
that the required number of bits/sample is smaller than that nominally required?
c). Assuming that effective compression can be achieved, would the proposed scheme
yield satisfactory results?
5.18 Solutions to Exercises in Chapter 5
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 5.2.1
For b-bit signed integers, the largest number is 2
b−1
− 1. For b = 32, we have
2,147,483,647 and for b = 64, we have 9,223,372,036,854,775,807 or about 9.2 × 10
Solution to Exercise 5.2.2
In floating point, the number of bits in the exponent determines the largest and
smallest representable numbers. For 32-bit floating point, the largest (smallest)
numbers are 2
±(127)
number is about 10
=1.7 × 10
9863
38
(5.9 × 10
−39
). For 64-bit floating point, the largest
.
Solution to Exercise 5.2.3
25 = 110112 and 7 = 1112. We find that 110012 + 1112 = 1000002 = 32.
Solution to Exercise 5.3.1
The only effect of pulse duration is to unequally weight the spectral repetitions.
Because we are only concerned with the repetition centered about the origin, the
pulse duration has no significant effect on recovering a signal from its samples.
Solution to Exercise 5.3.2
Figure 5.33
18
.
257
The square wave's spectrum is shown by the bolder set of lines centered about the
origin. The dashed lines correspond to the frequencies about which the spectral
repetitions (due to sampling with Ts = 1) occur. As the square wave's period decreases,
the negative frequency lines move to the left and the positive frequency ones to the
right.
Solution to Exercise 5.3.3
The simplest bandlimited signal is the sine wave. At the Nyquist frequency, exactly two
samples/period would occur. Reducing the sampling rate would result in fewer
samples/period, and these samples would appear to have arisen from a lower
frequency sinusoid.
Solution to Exercise 5.4.1
The plotted temperatures were quantized to the nearest degree. Thus, the high
temperature's amplitude was quantized as a form of A/D conversion.
Solution to Exercise 5.4.2
The signal-to-noise ratio does not depend on the signal amplitude. With an A/D range
of [−A, A], the quantization interval
and the signal's rms value (again assuming it is a sinusoid) is
Solution to Exercise 5.4.3
Solving 2
−B
= .001 results in B = 10 bits.
Solution to Exercise 5.4.4
A 16-bit A/D converter yields a SNR of 6 × 16 + 10log1.5 = 97.8 dB.
Solution to Exercise 5.6.1
(5.56)
Solution to Exercise 5.6.2
258
which, after manipulation, yields the geometric sum formula.
Solution to Exercise 5.6.3
If the sampling frequency exceeds the Nyquist frequency, the spectrum of the
samples equals the analog spectrum, but over the normalized analog frequency fT .
Thus, the energy in the sampled signal equals the original signal's energy multiplied by
T.
Solution to Exercise 5.7.1
This situation amounts to aliasing in the time-domain.
Solution to Exercise 5.8.1
When the signal is real-valued, we may only need half the spectral values, but the
complexity remains unchanged. If the data are complex-valued, which demands
retaining all frequency values, the complexity is again the same. When only
K frequencies are needed, the complexity is O (KN).
Solution to Exercise 5.9.1
If a DFT required 1ms to compute, and signal having ten times the duration would
require 100ms to compute. Using the FFT, a 1ms computing time would increase by a
factor of about 10log210 = 33 , a factor of 3 less than the DFT would have needed.
Solution to Exercise 5.9.2
The upper panel has not used the FFT algorithm to compute the length-4 DFTs while
the lower one has. The ordering is determined by the algorithm.
Solution to Exercise 5.9.3
The transform can have any greater than or equal to the actual duration of the signal.
We simply "pad" the signal with zero-valued samples until a computationally
advantageous signal length results. Recall that the FFT is an algorithm to compute the
DFT (Section 5.7). Extending the length of the signal this way merely means we are
sampling the frequency axis more fnely than required. To use the Cooley-Tukey
algorithm, the length of the resulting zero-padded signal can be 512, 1024, etc.
samples long.
Solution to Exercise 5.10.1
Number of samples equals 1.2 × 11025 = 13230. The datarate is 11025 × 16 = 176.4
kbps. The storage required would be 26460 bytes.
Solution to Exercise 5.10.2
The oscillations are due to the boxcar window's Fourier transform, which equals the
sinc function.
Solution to Exercise 5.10.3
These numbers are powers-of-two, and the FFT algorithm can be exploited with these
lengths. To compute a longer transform than the input signal's duration, we simply
zero-pad the signal.
259
Solution to Exercise 5.11.1
In discrete-time signal processing, an amplifier amounts to a multiplication, a very
easy operation to perform.
Solution to Exercise 5.12.1
The indices can be negative, and this condition is not allowed in MATLAB. To fix it, we
must start the signals later in the array.
Solution to Exercise 5.12.2
Such terms would require the system to know what future input or output values
would be before the current value was computed. Thus, such terms can cause
difficulties.
Solution to Exercise 5.13.1
It now acts like a bandpass filter with a center frequency of f0 and a bandwidth equal
to twice of the original lowpass filter.
Solution to Exercise 5.14.1
The DTFT of the unit sample equals a constant (equaling 1). Thus, the Fourier
transform of the output equals the transfer function.
Solution to Exercise 5.14.2
In sampling a discrete-time signal's Fourier transform L times equally over [0, 2π) to
form the DFT, the corresponding signal equals the periodic repetition of the original
signal.
(5.57)
To avoid aliasing (in the time domain), the transform length must equal or exceed the
signal's duration.
Solution to Exercise 5.14.3
The difference equation for an FIR filter has the form
(5.58)
The unit-sample response equals
(5.59)
260
which corresponds to the representation described in a problem (Discrete-Time
Systems in the Time-Domain (Page 228)) of a length-q boxcar filter.
Solution to Exercise 5.14.4
The unit-sample response's duration is q +1 and the signal's Nx. Thus the statement is
correct.
Solution to Exercise 5.15.1
Let N denote the input's total duration. The time-domain implementation requires a
total of N(2q + 1) computations, or 2q +1 computations per input value. In the
frequency domain, we split the input into
sections, each of which requires
per input in the section. Because we divide again by Nx to find the number of
computations per input value in the entire input, this quantity decreases as Nx
increases. For the time-domain implementation, it stays constant.
Solution to Exercise 5.15.2
The delay is not computational delay here the plot shows the first output value is
aligned with the filter's first input although in real systems this is an important
consideration. Rather, the delay is due to the filter's phase shift: A phase-shifted
sinusoid is equivalent to a time-delayed one:
All filters have phase shifts. This delay could be removed if the filter introduced no
phase shift. Such filters do not exist in analog form, but digital ones can be
programmed, but not in real time. Doing so would require the output to emerge
before the input arrives!
Solution to Exercise 5.16.1
We have p + q +1 multiplications and p + q − 1 additions. Thus, the total number of
arithmetic operations equals 2(p + q).
261
Chapter 6 Information
Communication
6.1 Information Communication
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
As far as a communications engineer is concerned, signals express information.
Because systems manipulate signals, they also affect the information content.
Information comes neatly packaged in both analog and digital forms. Speech, for
example, is clearly an analog signal, and computer files consist of a sequence of bytes,
a form of "discrete-time" signal despite the fact that the index sequences byte
position, not time sample. Communication systems endeavor not to manipulate
information, but to transmit it from one place to another, so-called point-to-point
communication, from one place to many others, broadcast communication, or
from many to many, like a telephone conference call or a chat room. Communication
systems can be fundamentally analog, like radio, or digital, like computer networks.
This chapter develops a common theory that underlies how such systems work. We
describe and analyze several such systems, some old like AM radio, some new like
computer networks. The question as to which is better, analog or digital
communication, has been answered, because of Claude Shannon's fundamental work
on a theory of information published in 1948, the development of cheap, highperformance computers, and the creation of high-bandwidth communication systems.
The answer is to use a digital communication strategy. In most cases, you should
convert all information-bearing signals into discrete-time, amplitude-quantized
signals. Fundamentally digital signals, like computer files (which are a special case of
symbolic signals), are in the proper form. Because of the Sampling Theorem, we know
how to convert analog signals into digital ones. Shannon showed that once in this
form, a properly engineered system can communicate digital information with
no error despite the fact that the communication channel thrusts noise onto all
transmissions. This startling result has no counterpart in analog systems; AM radio
will remain noisy. The convergence of these theoretical and engineering results on
communications systems has had important consequences in other arenas. The audio
compact disc (CD) and the digital videodisk (DVD) are now considered digital
communications systems, with communication design considerations used
throughout.
Go back to the fundamental model of communication (Figure 1.4). Communications
design begins with two fundamental considerations.
1. What is the nature of the information source, and to what extent can the receiver
tolerate errors in the received information?
2. What are the channel's characteristics and how do they afect the transmitted
signal?
262
In short, what are we going to send and how are we going to send it? Interestingly,
digital as well as analog transmission are accomplished using analog signals, like
voltages in Ethernet (an example of wireline communications) and electromagnetic
radiation (wireless) in cellular telephone.
6.2 Types of Communication Channels
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Electrical communications channels are either wireline or wireless channels. Wireline
channels physically connect transmitter to receiver with a "wire" which could be a
twisted pair, coaxial cable or optic fiber. Consequently, wireline channels are more
private and much less prone to interference. Simple wireline channels connect a single
transmitter to a single receiver: a point-to-point connection as with the telephone.
Listening in on a conversation requires that the wire be tapped and the voltage
measured. Some wireline channels operate in broadcast modes: one or more
transmitter is connected to several receivers. One simple example of this situation is
cable television. Computer networks can be found that operate in point-to-point or in
broadcast modes. Wireless channels are much more public, with a transmitter's
antenna radiating a signal that can be received by any antenna sufficiently close
enough. In contrast to wireline channels where the receiver takes in only the
transmitter's signal, the receiver's antenna will react to electromagnetic radiation
coming from any source. This feature has two faces: The smiley face says that a
receiver can take in transmissions from any source, letting receiver electronics select
wanted signals and disregarding others, thereby allowing portable transmission and
reception, while the frowny face says that interference and noise are much more
prevalent than in wireline situations. A noisier channel subject to interference
compromises the flexibility of wireless communication.
NOTE: You will hear the term tetherless networking applied to completely wireless
computer
networks.
Maxwell's equations neatly summarize the physics of all electromagnetic
phenomena, including cir cuits, radio, and optic fiber transmission.
(6.1)
where E is the electric field, H the magnetic field, ϵ dielectric permittivity, µ magnetic
permeability, σ electrical conductivity, and ρ is the charge density. Kirchof's Laws
represent special cases of these equations for circuits. We are not going to solve
263
Maxwell's equations here; do bear in mind that a fundamental understanding of
communications channels ultimately depends on fluency with Maxwell's equations.
Perhaps the most important aspect of them is that they are linear with respect to
the electrical and magnetic fields. Thus, the fields (and therefore the voltages and
currents) resulting from two or more sources will add.
Note: Nonlinear electromagnetic media do exist. The equations as written here are
simpler versions that apply to free-space propagation and conduction in metals.
Nonlinear media are becoming increasingly important in optic fiber communications,
which are also governed by Maxwell's equations.
6.3 Wireline Channels
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Wireline channels were the first used for electrical communications in the midnineteenth century for the telegraph. Here, the channel is one of several wires
connecting transmitter to receiver. The transmitter simply creates a voltage related to
the message signal and applies it to the wire(s). We must have a circuit a closed path
that supports current flow. In the case of single-wire communications, the earth is
used as the current's return path. In fact, the term ground for the reference node in
circuits originated in single-wire telegraphs. You can imagine that the earth's electrical
characteristics are highly variable, and they are. Single-wire metallic channels cannot
support high-quality signal transmission having a bandwidth beyond a few hundred
Hertz over any appreciable distance.
Figure 6.1 Coaxial Cable Cross-section
Coaxial cable consists of one conductor wrapped around the central conductor. This type of cable
supports broader bandwidth signals than twisted pair, and finds use in cable television and Ethernet.
Consequently, most wireline channels today essentially consist of pairs of conducting
wires (Figure 6.1 (Coaxial Cable Cross-section)), and the transmitter applies a messagerelated voltage across the pair. How these pairs of wires are physically configured
greatly affects their transmission characteristics. One example is twisted pair,
wherein the wires are wrapped about each other. Telephone cables are one example
of a twisted pair channel. Another is coaxial cable, where a concentric conductor
surrounds a central wire with a dielectric material in between. Coaxial cable, fondly
called "co-ax" by engineers, is what Ethernet uses as its channel. In either case,
wireline channels form a dedicated circuit between transmitter and receiver. As we
264
shall find subsequently, several transmissions can share the circuit by amplitude
modulation techniques; commercial cable TV is an example. These informationcarrying circuits are designed so that interference from nearby electromagnetic
sources is minimized. Thus, by the time signals arrive at the receiver, they are
relatively interference-and noise-free.
Both twisted pair and co-ax are examples of transmission lines, which all have the
circuit model shown in Figure 6.2 (Circuit Model for a Transmission Line) for an
infnitesimally small length. This circuit model arises from solving Maxwell's equations
for the particular transmission line geometry.
Figure 6.2 Circuit Model for a Transmission Line
The so-called distributed parameter model for two-wire cables has the depicted circuit model structure.
Element values depend on geometry and the properties of materials used to construct the transmission
line.
The series resistance comes from the conductor used in the wires and from the
conductor's geometry. The inductance and the capacitance derive from transmission
line geometry, and the parallel conductance from the medium between the wire pair.
Note that all the circuit elements have values expressed by the product of a constant
times a length; this notation represents that element values here have per-unit-length
units.
For example, the series resistance
has units of ohms/meter. For coaxial cable, the element values depend on the inner
conductor's radius ri, the outer radius of the dielectric rd, the conductivity of the
conductors σ, and the conductivity σd, dielectric constant ϵd, and magnetic permittivity
µd of the dielectric as
(6.2)
265
For twisted pair, having a separation d between the conductors that have conductivity
σ and common radius r and that are immersed in a medium having dielectric and
magnetic properties, the element values are then
(6.3)
The voltage between the two conductors and the current flowing through them will
depend on distance x along the transmission line as well as time. We express this
dependence as v (x , t) and i (x , t). When we place a sinusoidal source at one end of the
transmission line, these voltages and currents will also be sinusoidal because the
transmission line model consists of linear circuit elements. As is customary in
analyzing linear circuits, we express voltages and currents as the real part of complex
exponential signals, and write circuit variables as a complex amplitude here
dependent on distance times a complex exponential: v (x , t)= Re (V (x) e
j2πft
) and i (x ,
t)= Re (I (x) e j2πft). Using the transmission line circuit model, we find from KCL, KVL, and
v-i relations the equations governing the complex amplitudes.
KCL at Center Node
(6.4)
V-I relation for RL series
(6.5)
Rearranging and taking the limit Δx → 0 yields the so-called transmission line
equations.
(6.6)
By combining these equations, we can obtain a single equation that governs how the
voltage's or the current's complex amplitude changes with position along the
266
transmission line. Taking the derivative of the second equation and plugging the first
equation into the result yields the equation governing the voltage.
(6.7)
This equation's solution is
(6.8)
Calculating its second derivative and comparing the result with our equation for the
voltage can check this solution.
(6.9)
Our solution works so long as the quantity γ satisfies
(6.10)
Thus, γ depends on frequency, and we express it in terms of real and imaginary parts
as indicated. The quantities V+ and V− are constants determined by the source and
physical considerations. For example, let the spatial origin be the middle of the
transmission line model F igure 6.2 (Circuit Model for a Transmission Line). Because
the circuit model contains simple circuit elements, physically possible solutions for
voltage amplitude cannot increase with distance along the transmission line.
Expressing γ in terms of its real and imaginary parts in our solution shows that such
increases are a (mathematical) possibility.
The voltage cannot increase without limit; because a (f) is always positive, we must
segregate the solution for negative and positive x. The first term will increase
exponentially for x < 0 unless V+ = 0 in this region; a similar result applies to V− for x >
0. These physical constraints give us a cleaner solution.
(6.11)
This solution suggests that voltages (and currents too) will decrease exponentially
along a transmission line. The space constant, also known as the attenuation
constant, is the distance over which the voltage decreases by a factor of
267
It equals the reciprocal of a(f), which depends on frequency, and is expressed by
manufacturers in units of dB/m.
The presence of the imaginary part of γ, b(f), also provides insight into how
−(jbx)
transmission lines work. Because the solution for x > 0 is proportional to e
, we
know that the voltage's complex amplitude will vary sinusoidally in space. The
complete solution for the voltage has the form
(6.12)
The complex exponential portion has the form of a propagating wave. If we could
take a snapshot of the voltage (take its picture at t = t1), we would see a sinusoidally
varying waveform along the transmission line. One period of this variation, known as
the wavelength, equals
If we were to take a second picture at some later time t = t2, we would also see a
sinusoidal voltage. Because
the second waveform appears to be the first one, but delayed shifted to the right in
space. Thus, the voltage appeared to move to the right with a speed equal to
(assuming b > 0). We denote this propagation speed by c, and it equals 2πf
(6.13)
In the high-frequency region where
and
the quantity under the radical simplifies to
and we find the propagation speed to be
(6.14)
268
For typical coaxial cable, this propagation speed is a fraction (one-third to two-thirds)
of the speed of light.
Exercise 6.3.1
Find the propagation speed in terms of physical parameters for both the coaxial cable
and twisted pair examples.
By using the second of the transmission line equation (6.6), we can solve for the
current's complex amplitude. Considering the spatial region x > 0, for example, we find
that
which means that the ratio of voltage and current complex amplitudes does not
depend on distance.
(6.15)
The quantity Z0 is known as the transmission line's characteristic impedance. Note
that when the signal frequency is sufciently high, the characteristic impedance is real,
which means the transmission line appears resistive in this high-frequency regime.
(6.16)
Typical values for characteristic impedance are 50 and 75 Ω.
A related transmission line is the optic fiber. Here, the electromagnetic field is light,
and it propagates down a cylinder of glass. In this situation, we don't have two
conductors in fact we have none and the energy is propagating in what corresponds
to the dielectric material of the coaxial cable. Optic fiber com munication has exactly
the same properties as other transmission lines: Signal strength decays exponentially
according to the fiber's space constant and propagates at some speed less than light
would in free space. From the encompassing view of Maxwell's equations, the only
difference is the electromagnetic signal's frequency. Because no electric conductors
are present and the fiber is protected by an opaque "insulator," optic fiber
transmission is interference-free.
Exercise 6.3.2
From tables of physical constants, find the frequency of a sinusoid in the middle of the
visible light range. Compare this frequency with that of a mid-frequency cable television
signal. To summarize, we use transmission lines for high-frequency wireline signal
communication. In wireline communication, we have a direct, physical connection a circuit
between transmitter and receiver. When we select the transmission line characteristics and
269
the transmission frequency so that we operate in the high-frequency regime, signals are
not filtered as they propagate along the transmission line: The characteristic impedance is
real-valued the transmission line's equivalent impedance is a resistor and all the signal's
components at various frequencies propagate at the same speed. Transmitted signal
amplitude does decay exponentially along the transmission line. Note that in the highfrequency regime the space constant is approximately zero, which means the attenuation
is quite small.
Exercise 6.3.3
What is the limiting value of the space constant in the high frequency regime?
6.4 Wireless Channels
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Wireless channels exploit the prediction made by Maxwell's equation that
electromagnetic fields propagate in free space like light. When a voltage is applied to
an antenna, it creates an electromagnetic field that propagates in all directions
(although antenna geometry affects how much power flows in any given direction)
that induces electric currents in the receiver's antenna. Antenna geometry determines
how energetic a field a voltage of a given frequency creates. In general terms, the
dominant factor is the relation of the antenna's size to the field's wavelength. The
fundamental equation relating frequency and wavelength for a propagating wave is
λf = c
Thus, wavelength and frequency are inversely related: High frequency corresponds to
small wavelengths. For example, a 1 MHz electromagnetic field has a wavelength of
300 m. Antennas having a size or distance from the ground comparable to the
wavelength radiate fields most efficiently. Consequently, the lower the frequency the
bigger the antenna must be. Because most information signals are baseband signals,
having spectral energy at low frequencies, they must be modulated to higher
frequencies to be transmitted over wireless channels.
For most antenna-based wireless systems, how the signal diminishes as the receiver
moves further from the transmitter derives by considering how radiated power
changes with distance from the transmitting antenna. An antenna radiates a given
amount of power into free space, and ideally this power propagates without loss in all
directions. Considering a sphere centered at the transmitter, the total power, which is
found by integrating the radiated power over the surface of the sphere, must be
constant regardless of the sphere's radius. This requirement results from the
conservation of energy. Thus, if p (d) represents the power integrated with respect to
direction at a distance d from the antenna, the total power will be p (d)4πd2 . For this
quantity to be a constant, we must have
270
which means that the received signal amplitude AR must be proportional to the
transmitter's amplitude AT and inversely related to distance from the transmitter.
(6.17)
for some value of the constant k. Thus, the further from the transmitter the receiver is
located, the weaker the received signal. Whereas the attenuation found in wireline
channels can be controlled by physical parameters and choice of transmission
frequency, the inverse-distance attenuation found in wireless channels persists across
all frequencies.
Exercise 6.4.1
Why don't signals attenuate according to the inverse-square law in a conductor? What
is the difference between the wireline and wireless cases?
The speed of propagation is governed by the dielectric constant µ0 and magnetic
permeability E0 of free space.
(6.18)
Known familiarly as the speed of light, it sets an upper limit on how fast signals can
propagate from one place to another. Because signals travel at a finite speed, a
receiver senses a transmitted signal only after a time delay inversely related to the
propagation speed:
(6.19)
At the speed of light, a signal travels across the United States in 16 ms, a reasonably
small time delay. If a lossless (zero space constant) coaxial cable connected the East
and West coasts, this delay would be two to three times longer because of the slower
propagation speed.
6.5 Line-of-Sight Transmission
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Long-distance transmission over either kind of channel encounters attenuation
problems. Losses in wireline channels are explored in the Circuit Models module
(Wireline Channels (Page 263)), where repeaters can extend the distance between
transmitter and receiver beyond what passive losses the wireline channel imposes. In
wireless channels, not only does radiation loss occur, but also one antenna may not
"see" another because of the earth's curvature.
271
Figure 6.3
Two antennae are shown each having the same height. Line-of-sight transmission means the transmitting
and receiving antennae can "see" each other as shown. The maximum distance at which they can see
each other, dLOS, occurs when the sighting line just grazes the earth's surface.
At the usual radio frequencies, propagating electromagnetic energy does not follow
the earth's surface. Line-of-sight communication has the transmitter and receiver
antennas in visual contact with each other. Assuming both antennas have height h
above the earth's surface, maximum line-of-sight distance is
(6.19)
6
where R is the earth's radius ( 6.38 × 10 m).
Exercise 6.5.1
Derive the expression of line-of-sight distance using only the Pythagorean Theorem.
Generalize it to the case where the antennas have different heights (as is the case with
commercial radio and cellular telephone). What is the range of cellular telephone
where the handset antenna has essentially zero height?
Exercise 6.5.2
Can you imagine a situation wherein global wireless communication is possible with
only one transmitting antenna? In particular, what happens to wavelength when
carrier frequency decreases?
Using a 100 m antenna would provide line-of-sight transmission over a distance of
71.4 km. Using such very tall antennas would provide wireless communication within a
town or between closely spaced population centers. Consequently, networks of
antennas sprinkle the countryside (each located on the highest hill possible) to
provide long-distance wireless communications: Each antenna receives energy from
one antenna and retransmits to another. This kind of network is known as a relay
network.
272
6.6 The Ionosphere and Communications
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
If we were limited to line-of-sight communications, long distance wireless
communication, like ship-to-shore communication, would be impossible. At the turn of
the century, Marconi, the inventor of wireless telegraphy, boldly tried such long
distance communication without any evidence either empirical or theoretical that it
was possible. When the experiment worked, but only at night, physicists scrambled to
determine why (using Maxwell's equations, of course). It was Oliver Heaviside, a
mathematical physicist with strong engineering interests, who hypothesized that an
invisible electromagnetic "mirror" surrounded the earth.
What he meant was that at optical frequencies (and others as it turned out), the mirror
was transparent, but at the frequencies Marconi used, it reflected electromagnetic
radiation back to earth. He had predicted the existence of the ionosphere, a plasma
that encompasses the earth at altitudes hi between 80 and 180 km that reacts to solar
radiation: It becomes transparent at Marconi's frequencies during the day, but
becomes a mirror at night when solar radiation diminishes. The maximum distance
along the earth's surface that can be reached by a single ionospheric refection is
which ranges between 2,010 and 3,000 km when we substitute minimum and
maximum ionospheric altitudes. This distance does not span the United States or
cross the Atlantic; for transatlantic communication, at least two reflections would be
required.
The communication delay encountered with a single refection in this channel is
which ranges between 6.8 and 10 ms, again a small time interval.
6.7 Communication with Satellites
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Global wireless communication relies on satellites. Here, ground stations transmit to
orbiting satellites that amplify the signal and retransmit it back to earth. Satellites will
move across the sky unless they are in geosynchronous orbits, where the time for
one revolution about the equator exactly matches the earth's rotation time of one
day. TV satellites would require the homeowner to continually adjust his or her
antenna if the satellite weren't in geosynchronous orbit. Newton's equations applied
to orbiting bodies predict that the time T for one orbit is related to distance from the
earth's center R as
273
(6.20)
where G is the gravitational constant and M the earth's mass. Calculations yield R =
42200km, which corresponds to an altitude of 35700km. This altitude greatly exceeds
that of the ionosphere, requiring satellite transmitters to use frequencies that pass
through it. Of great importance in satellite communications is the transmission delay.
The time for electromagnetic fields to propagate to a geosynchronous satellite and
return is 0.24 s, a significant delay.
Exercise 6.7.1
In addition to delay, the propagation attenuation encountered in satellite
communication far exceeds what occurs in ionospheric-mirror based communication.
Calculate the attenuation incurred by radiation going to the satellite (one-way loss)
with that encountered by Marconi (total going up and down). Note that the
attenuation calculation in the ionospheric case, assuming the ionosphere acts like a
perfect mirror, is not a straightforward application of the propagation loss formula.
6.8 Noise and Interference
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We have mentioned that communications are, to varying degrees, subject to
interference and noise. It's time to be more precise about what these quantities are
and how they difer.
Interference represents man-made signals. Telephone lines are subject to power-line
interference (in the United States a distorted 60 Hz sinusoid). Cellular telephone
channels are subject to adjacent-cell phone conversations using the same signal
frequency. The problem with such interference is that it occupies the same frequency
band as the desired communication signal, and has a similar structure.
Exercise 6.8.1
Suppose interference occupied a different frequency band; how would the receiver
remove it?
We use the notation i (t) to represent interference. Because interference has manmade structure, we can write an explicit expression for it that may contain some
unknown aspects (how large it is, for example).
Noise signals have little structure and arise from both human and natural sources.
Satellite channels are subject to deep space noise arising from electromagnetic
radiation pervasive in the galaxy. Thermal noise plagues all electronic circuits that
contain resistors. Thus, in receiving small amplitude signals, receiver amplifiers will
most certainly add noise as they boost the signal's amplitude. All channels are subject
to noise, and we need a way of describing such signals despite the fact we can't write a
formula for the noise signal like we can for interference. The most widely used noise
model is white noise. It is defined entirely by its frequency-domain characteristics.
• White noise has constant power at all frequencies.
274
• At each frequency, the phase of the noise spectrum is totally uncertain: It can be
any value in between 0 and 2π, and its value at any frequency is unrelated to the
phase at any other frequency.
• When noise signals arising from two different sources add, the resultant noise
signal has a power equal to the sum of the component powers.
Because of the emphasis here on frequency-domain power, we are led to define the
power spectrum. Because of Parseval's Theorem9, we define the power spectrum Ps
(f) of a non-noise signal s (t) to be the magnitude-squared of its Fourier transform.
(6.21)
Integrating the power spectrum over any range of frequencies equals the power the
signal contains in that band. Because signals must have negative frequency
components that mirror positive frequency ones, we routinely calculate the power in a
spectral band as the integral over positive frequencies multiplied by two.
(6.22)
Using the notation n (t) to represent a noise signal's waveform, we define noise in
terms of its power spectrum. For white noise, the power spectrum equals the
constant
With this definition, the power in a frequency band equals N0 (f2 -f1).
When we pass a signal through a linear, time-invariant system, the output's spectrum
equals the product of the system's frequency response and the input's spectrum.
Thus, the power spectrum of the system's output is given by
(6.23)
This result applies to noise signals as well. When we pass white noise through a filter,
the output is also a noise signal but with power spectrum
6.9 Channel Models
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Both wireline and wireless channels share characteristics, allowing us to use a
common model for how the channel affects transmitted signals.
• The transmitted signal is usually not filtered by the channel.
275
• The signal can be attenuated.
• The signal propagates through the channel at a speed equal to or less than the
speed of light, which means that the channel delays the transmission.
• The channel may introduce additive interference and/or noise.
Letting α represent the attenuation introduced by the channel, the receiver's input
signal is related to the transmitted one by
(6.24)
This expression corresponds to the system model for the channel shown in Figure
6.4 In this book, we shall assume that the noise is white.
Figure 6.4
The channel component of the fundamental model of communication (Figure 1.4: Fundamental model of
communication) has the depicted form. The attenuation is due to propagation loss. Adding the
interference and noise is justifed by the linearity property of Maxwell's equations.
Exercise 6.9.1
Is this model for the channel linear? As expected, the signal that emerges from the
channel is corrupted, but does contain the transmitted signal. Communication system
design begins with detailing the channel model, then developing the transmitter and
receiver that best compensate for the channel's corrupting behavior. We characterize
the channel's quality by the signal-to-interference ratio (SIR) and the signal-to-noise
ratio (SNR). The ratios are computed according to the relative power of each within
the transmitted signal's bandwidth. Assuming the signal x (t)'s spectrum spans the
frequency interval [fl,fu], these ratios can be expressed in terms of power spectra.
(6.25)
(6.26)
In most cases, the interference and noise powers do not vary for a given receiver.
Variations in signal-tointerference and signal-to-noise ratios arise from the
attenuation because of transmitter-to-receiver distance variations.
276
6.10 Baseband Communication
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We use analog communication techniques for analog message signals, like music,
speech, and television. Transmission and reception of analog signals using analog
results in an inherently noisy received signal (assuming the channel adds noise, which
it almost certainly does).
The simplest form of analog communication is baseband communication.
POINT OF INTEREST: We use analog communication techniques for analog message
signals, like
music, speech, and television. Transmission and reception of analog signals using
analog results in an inherently noisy received signal (assuming the channel adds noise,
which it almost certainly does).
Here, the transmitted signal equals the message times a transmitter gain.
(6.27)
An example, which is somewhat out of date, is the wireline telephone system. You
don't use baseband communication in wireless systems simply because low-frequency
signals do not radiate well. The receiver in a baseband system can't do much more
than filter the received signal to remove out-of-band noise (interference is small in
wireline channels). Assuming the signal occupies a bandwidth of W Hz (the signal's
spectrum extends from zero to W) , the receiver applies a lowpass filter having the
same bandwidth, as shown in Figure 6.5.
Figure 6.5
The receiver for baseband communication systems is quite simple: a lowpass filter having the same
bandwidth as the signal.
We use the signal-to-noise ratio of the receiver's output m (t) to evaluate any analogmessage com munication system. Assume that the channel introduces an attenuation
α and white noise of spectral height
The filter does not affect the signal component we assume its gain is unity but does
filter the noise, removing frequency components above W Hz. In the filter's output, the
2 2
receiv ed signal power equals α G power (m) and the noise power N0W, which gives a
signal-to-noise ratio of
277
(6.28)
The signal power (m) will be proportional to the bandwidth W; thus, in baseband
communication the signal-to-noise ratio varies only with transmitter gain and channel
attenuation and noise level.
6.11 Modulated Communication
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Especially for wireless channels, like commercial radio and television, but also for
wireline systems like cable television, an analog message signal must be modulated:
The transmitted signal's spectrum occurs at much higher frequencies than those
occupied by the signal.
POINT OF INTEREST: We use analog communication techniques for analog message
signals, like music, speech, and television. Transmission and reception of analog
signals using analog results in an inherently noisy received signal (assuming the
channel adds noise, which it almost certainly does).
The key idea of modulation is to affect the amplitude, frequency or phase of what is
known as the carrier sinusoid. Frequency modulation (FM) and less frequently used
phase modulation (PM) are not discussed here; we focus on amplitude modulation
(AM). The amplitude modulated message signal has the form
(6.29)
where fc is the carrier frequency and Ac the carrier amplitude. Also, the signal's
amplitude is assumed to be less than one: |m (t) | < 1. From our previous exposure to
amplitude modulation (see the Fourier Transform example (Example 4.5)), we know
that the transmitted signal's spectrum occupies the frequency range [fc − W, fc + W ],
assuming the signal's bandwidth is W Hz (see the figure (Figure 6.6)). The carrier
frequency is usually much larger than the signal's highest frequency: fc » W , which
means that the transmitter antenna and carrier frequency are chosen jointly during
the design process.
Figure 6.6
278
The AM coherent receiver along with the spectra of key signals is shown for the case of a triangularshaped signal spectrum. The dashed line indicates the white noise level. Note that the filters'
characteristics cutoff frequency and center frequency for the bandpass filter must be match to the
modulation and message parameters.
Ignoring the attenuation and noise introduced by the channel for the moment,
reception of an amplitude modulated signal is quite easy (see Problem 4.20). The socalled coherent receiver multiplies the input signal by a sinusoid and lowpass-filters
the result (Figure 6.6).
(6.30)
Because of our trigonometric identities, we know that
(6.31)
At this point, the message signal is multiplied by a constant and a sinusoid at twice the
carrier frequency. Multiplication by the constant term returns the message signal to
baseband (where we want it to be!) while multiplication by the double-frequency term
yields a very high frequency signal. The lowpass filter removes this high-frequency
signal, leaving only the baseband signal. Thus, the received signal is
(6.32)
Exercise 6.11.1
This derivation relies solely on the time domain; derive the same result in the
frequency domain. You won't need the trigonometric identity with this approach.
Because it is so easy to remove the constant term by electrical means we insert a
capacitor in series with the receiver's output we typically ignore it and concentrate on
the signal portion of the receiver's output when calculating signal-to-noise ratio.
6.12 Signal-to-Noise Ratio of an Amplitude-Modulated
Signal
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we consider the much more realistic situation when we have a channel that
introduces attenuation and noise, we can make use of the just-described receiver's
linear nature to directly derive the receiver's output. The attenuation affects the
output in the same way as the transmitted signal: It scales the output signal by the
same amount. The white noise, on the other hand, should be filtered from the
received signal before demodulation. We must thus insert a bandpass filter having
279
bandwidth 2W and center frequency fc: This filter has no effect on the received signalrelated component, but does remove out-of-band noise power. As shown in the
triangular-shaped signal spectrum (Figure 6.6), we apply coherent receiver to this
filtered signal, with the result that the demodulated output contains noise that cannot
be removed: It lies in the same spectral band as the signal.
As we derive the signal-to-noise ratio in the demodulated signal, let's also calculate the
signal-to-noise ratio of the bandpass filter's output
The signal component of
equals αAcm (t) cos (2πfct) . This signal's Fourier transform equals
(6.33)
making the power spectrum,
(6.34)
Exercise 6.12.1
If you calculate the magnitude-squared of the first equation, you don't obtain the
second unless you make an assumption. What is it?
Thus, the total signal-related power in
is
The noise power equals the integral of the noise power spectrum; because the power
spectrum is constant over the transmission band, this integral equals the noise
amplitude N0 times the filter's bandwidth 2W . The so-called received signal-to-noise
ratio the signal-to-noise ratio after the de rigeur front-end bandpass filter and before
demodulation equals
(6.35)
The demodulated signal
Clearly, the signal power equals
To determine the noise power, we must understand how the coherent demodulator
afects the bandpass noise found in
280
Because we are concerned with noise, we must deal with the power spectrum since
we don't have the Fourier transform available to us. Letting P (f) denote the power
spectrum of
's noise component, the power spectrum after multiplication by the carrier has the
form
(6.36)
The delay and advance in frequency indicated here results in two spectral noise bands
falling in the low-frequency region of lowpass filter's passband. Thus, the total noise
power in this filter's output equals
The signal-to-noise ratio of the receiver's output thus equals
(6.37)
Let's break down the components of this signal-to-noise ratio to better appreciate
how the channel and the transmitter parameters affect communications
performance. Better performance, as measured by the SNR, occurs as it increases.
• More transmitter power increasing AC increases the signal-to-noise ratio
proportionally.
• The carrier frequency fc has no effect on SNR, but we have assumed that fc » W .
• The signal bandwidth W enters the signal-to-noise expression in two places:
implicitly through the signal power and explicitly in the expression's denominator.
If the signal spectrum had a constant amplitude as we increased the
bandwidth, signal power would increase proportionally. On the other hand, our
transmitter enforced the criterion that signal amplitude was constant
(Communication with Satellites (Page 272)). Signal amplitude essentially equals
the integral of the magnitude of the signal's spectrum.
NOTE: This result isn't exact, but we do know that
Enforcing the signal amplitude specification means that as the signal's bandwidth
increases we must decrease the spectral amplitude, with the result that the signal
power remains constant. Thus, increasing signal bandwidth does indeed decrease the
signal-to-noise ratio of the receiver's output.
• Increasing channel attenuation moving the receiver farther from the transmitter
decreases the signal-to-noise ratio as the square. Thus, signal-to-noise ratio
decreases as distance-squared between transmitter and receiver.
281
• Noise added by the channel adversely affects the signal-to-noise ratio.
In summary, amplitude modulation provides an effective means for sending a
bandlimited signal from one place to another. For wireline channels, using baseband
or amplitude modulation makes little difference in terms of signal-to-noise ratio. For
wireless channels, amplitude modulation is the only alternative. The one AM
parameter that does not affect signal-to-noise ratio is the carrier frequency fc: We can
choose any value we want so long as the transmitter and receiver use the same value.
However, suppose someone else wants to use AM and chooses the same carrier
frequency. The two resulting transmissions will add, and both receivers will produce
the sum of the two signals. What we clearly need to do is talk to the other party, and
agree to use separate carrier frequencies. As more and more users wish to use radio,
we need a forum for agreeing on carrier frequencies and on signal bandwidth. On
earth, this forum is the government. In the United States, the Federal Communications
Commission (FCC) strictly controls the use of the electromagnetic spectrum for
communications. Separate frequency bands are allocated for commercial AM, FM,
cellular telephone (the analog version of which is AM), short wave (also AM), and
satellite communications.
Exercise 6.12.2
Suppose all users agree to use the same signal bandwidth. How closely can the carrier
frequencies be while avoiding communications crosstalk? What is the signal
bandwidth for commercial AM? How does this bandwidth compare to the speech
bandwidth?
6.13 Digital Communication
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Effective, error-free transmission of a sequence of bits --- a bit stream {b (0) ,b (1) ,... }
is the goal here. We found that analog schemes, as represented by amplitude
modulation, always yield a received signal containing noise as well as the message
signal when the channel adds noise. Digital communication schemes are very
different. Once we decide how to represent bits by analog signals that can be
transmitted over wireline (like a computer network) or wireless (like digital cellular
telephone) channels, we will then develop a way of tacking on communication bits to
the message bits that will reduce channel-induced errors greatly. In theory, digital
communication errors can be zero, even though the channel adds noise!
We represent a bit by associating one of two specific analog signals with the bit's
value. Thus, if b (n)=0, we transmit the signal s0 (t); if b (n)=1, send s1 (t). These two
signals comprise the signal set for digital communication and are designed with the
channel and bit stream in mind. In virtually every case, these signals have a finite
duration T common to both signals; this duration is known as the bit interval. Exactly
what signals we use ultimately affects how well the bits can be received. Interestingly,
baseband and modulated signal sets can yield the same performance. Other
considerations determine how signal set choice afects digital communication
performance.
282
Exercise 6.13.1
What is the expression for the signal arising from a digital transmitter sending the bit
stream b (n), n = {..., −1, 0, 1,... } using the signal set s 0 (t), s1 (t), each signal of which
has duration T ?
6.14 Binary Phase Shift Keying
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
A commonly used example of a signal set consists of pulses that are negatives of each
other (Figure 6.7).
(6.38)
Figure 6.7
Here, we have a baseband signal set suitable for wireline transmission. The entire bit
stream b (n) is represented by a sequence of these signals. Mathematically, the
transmitted signal has the form
(6.39)
and graphically Figure 6.8 shows what a typical transmitted signal might be.
283
Figure 6.8
The upper plot shows how a baseband signal set for transmitting the bit sequence 0110. The lower one
shows an amplitude-modulated variant suitable for wireless channels.
This way of representing a bit stream changing the bit changes the sign of the
transmitted signal is known as binary phase shift keying and abbreviated BPSK. The
name comes from concisely expressing this popular way of communicating digital
information. The word "binary" is clear enough (one binary-valued quantity is
transmitted during a bit interval). Changing the sign of sinusoid amounts to changing
shifting the phase by π (although we don't have a sinusoid yet). The word "keying"
reflects back to the first electrical communication system, which happened to be
digital as well: the telegraph.
The datarate R of a digital communication system is how frequently an information
bit is transmitted. In this example it equals the reciprocal of the bit interval :
Thus, for a 1 Mbps (megabit per second) transmission, we must have T =1µs .
The choice of signals to represent bit values is arbitrary to some degree. Clearly, we do
not want to choose signal set members to be the same; we couldn't distinguish bits if
we did so. We could also have made the negative-amplitude pulse represent a 0 and
the positive one a 1. This choice is indeed arbitrary and will have no effect on
performance assuming the receiver knows which signal represents which bit. As in all
communication systems, we design transmitter and receiver together.
A simple signal set for both wireless and wireline channels amounts to amplitude
modulating a baseband signal set (more appropriate for a wireline channel) by a
carrier having a frequency harmonic with the bit interval
284
(6.40)
Figure 6.9
Exercise 6.14.1
What is the value of k in this example?
This signal set is also known as a BPSK signal set. We'll show later that indeed both
signal sets provide identical performance levels when the signal-to-noise ratios are
equal.
Exercise 6.14.2
Write a formula, in the style of the baseband signal set, for the transmitted signal as
shown in the plot of the baseband signal set that emerges when we use this
modulated signal.
What is the transmission bandwidth of these signal sets? We need only consider the
baseband version as the second is an amplitude-modulated version of the first. The
bandwidth is determined by the bit sequence. If the bit sequence is constant always 0
or always 1 the transmitted signal is a constant, which has zero bandwidth. The worstcase bandwidth consuming bit sequence is the alternating one shown in Figure 6.10.
In this case, the transmitted signal is a square wave having a period of 2T .
Figure 6.10
Here we show the transmitted waveform corresponding to an alternating bit sequence.
From our work in Fourier series, we know that this signal's spectrum contains oddharmonics of the fundamental, which here equals
Thus, strictly speaking, the signal's bandwidth is infinite. In practical terms, we use the
90%-power bandwidth to assess the effective range of frequencies consumed by the
signal. The first and third harmonics contain that fraction of the total power, meaning
that the effective bandwidth of our baseband signal is
285
or, expressing this quantity in terms of the datarate,
Thus, a digital communications signal requires more bandwidth than the datarate: a 1
Mbps baseband system requires a bandwidth of at least 1.5 MHz. Listen carefully
when someone describes the transmission bandwidth of digital communication
systems: Did they say "megabits" or "megahertz"?
Exercise 6.14.3
Show that indeed the first and third harmonics contain 90% of the transmitted power.
If the receiver uses a front-end filter of bandwidth
what is the total harmonic distortion of the received signal?
Exercise 6.14.4
What is the 90% transmission bandwidth of the modulated signal set?
6.15 Frequency Shift Keying
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
In frequency-shift keying (FSK), the bit affects the frequency of a carrier sinusoid.
(6.41)
Figure 6.11
The frequencies f0, f1 are usually harmonically related to the bit interval. In the
depicted example,
and
As can be seen from the transmitted signal for our example bit stream (Figure 6.12),
the transitions at bit interval boundaries are smoother than those of BPSK.
Figure 6.12
286
This plot shows the FSK waveform for same bitstream used in the BPSK example (Figure 6.8).
To determine the bandwidth required by this signal set, we again consider the
alternating bit stream. Think of it as two signals added together: The first comprised of
the signal s0 (t), the zero signal, s0(t), zero, etc., and the second having the same
structure but interleaved with the first and containing s1 (t) (Figure 6.13).
Figure 6.13
The depicted decomposition of the FSK-modulated alternating bit stream into its frequency components
simplifies the calculation of its bandwidth.
Each component can be thought of as a fixed-frequency sinusoid multiplied by a
square wave of period 2T that alternates between one and zero. This baseband
square wave has the same Fourier spectrum as our BPSK example, but with the
addition of the constant term c0. This quantity's presence changes the number of
Fourier series terms required for the 90% bandwidth: Now we need only include the
zero and first harmonics to achieve it. The bandwidth thus equals, with
If the two frequencies are harmonics of the bit-interval duration,
and
with k1 > k0, the bandwidth equals
If the difference between harmonic numbers is 1, then the FSK bandwidth is smaller
than the BPSK bandwidth. If the difference is 2, the bandwidths are equal and larger
differences produce a transmission bandwidth larger than that resulting from using a
BPSK signal set.
287
6.16 Digital Communication Receivers
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The receiver interested in the transmitted bit stream must perform two tasks when
received waveform r (t) begins.
• It must determine when bit boundaries occur: The receiver needs to synchronize
with the transmitted signal. Because transmitter and receiver are designed in
concert, both use the same value for the bit interval T. Synchronization can occur
because the transmitter begins sending with a reference bit sequence, known as
the preamble. This reference bit sequence is usually the alternating sequence as
shown in the square wave example and in the FSK example (Figure 6.13). The
receiver knows what the preamble bit sequence is and uses it to determine when
bit boundaries occur. This procedure amounts to what in digital hardware as selfclockingsignaling: The receiver of a bit stream must derive the clock when bit
boundaries occur from its input signal. Because the receiver usually does not
determine which bit was sent until synchronization occurs, it does not know when
during the preamble it obtained synchronization. The transmitter signals the end
of the preamble by switching to a second bit sequence. The second preamble
phase informs the receiver that data bits are about to come and that the
preamble is almost over.
• Once synchronized and data bits are transmitted, the receiver must then
determine every T seconds what bit was transmitted during the previous bit
interval. We focus on this aspect of the digital receiver because this strategy is
also used in synchronization. The receiver for digital communication is known as
a matched filter. Optimal receiver structure
Figure 6.14 Optimal receiver structure
The optimal receiver structure for digital communication faced with additive white noise channels is the
depicted matched filter.
This receiver, shown in Figure 6.14 (Optimal receiver structure), multiplies the received
signal by each of the possible members of the transmitter signal set, integrates the
product over the bit interval, and compares the results. Whichever path through the
receiver yields the largest value corresponds to the receiver's decision as to what bit
was sent during the previous bit interval. For the next bit interval, the multiplication
and integration begins again, with the next bit decision made at the end of the bit
interval. Mathematically, the received value of b (n), which we label
288
is given by
(6.42)
You may not have seen the
notation before. maxi{i, ·} yields the maximum value of its argument with respect to
the index
equals the value of the index that yields the maximum. Note that the precise
numerical value of the integrator's output does not matter; what does matter is its
value relative to the other integrator's output.
Let's assume a perfect channel for the moment: The received signal equals the
transmitted one. If bit 0 were sent using the baseband BPSK signal set, the integrator
outputs would be
(6.43)
If bit 1 were sent,
(6.44)
Exercise 6.16.1
Can you develop a receiver for BPSK signal sets that requires only one multiplierintegrator combination?
Exercise 6.16.2
What is the corresponding result when the amplitude-modulated BPSK signal set is
used? Clearly, this receiver would always choose the bit correctly. Channel attenuation
would not affect this correctness; it would only make the values smaller, but all that
matters is which is largest.
289
6.17 Digital Communication in the Presence of Noise
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
When we incorporate additive noise into our channel model, so that r(t)=αsi(t)+n(t),
errors can creep in. If the transmitter sent bit 0 using a BPSK signal set (Section 6.14),
the integrators' outputs in the matched filter receiver (Figure 6.14: Optimal receiver
structure) would be:
(6.45)
It is the quantities containing the noise terms that cause errors in the receiver's
decision-making process. Because they involve noise, the values of these integrals are
random quantities drawn from some probability distribution that vary erratically from
bit interval to bit interval. Because the noise has zero average value and has an equal
amount of power in all frequency bands, the values of the integrals will hover about
zero. What is important is how much they vary. If the noise is such that its integral
2
term is more negative than αA T, then the receiver will make an error, deciding that
the transmitted zero-valued bit was indeed a one. The probability that this situation
occurs depends on three factors:
• Signal Set Choice - The difference between the signal-dependent terms in the
integrators' outputs (equations (6.45)) defines how large the noise term must be
for an incorrect receiver decision to result. What affects the probability of such
errors occurring is the square of this difference in comparison tothe noise term's
2 4 2
variability. For our BPSK baseband signal set, the signal-related value is 4α A T .
• Variability of the Noise Term - We quantify variability by the average value of its
square, which is essentially the noise term's power. This calculation is best
performed in the frequency domain and equals
• Because of Parseval's Theorem, we know that
2
which for the baseband signal set equals A T. Thus, the noise term's power is
• Probability Distribution of the Noise Term - The value of the noise terms
relative to the signal terms and the probability of their occurrence directly affect
290
the likelihood that a receiver error will occur. For the white noise we have been
considering, the underlying distributions are Gaussian. The probability the
receiver makes an error on any bit transmission equals:
Here Q (·) is the integral
This integral has no closed form expression, but it can be accurately computed. As
Figure 6.15 illustrates, Q (·) is a decreasing, very nonlinear function.
Figure 6.15
The function Q(x) is plotted in semilogarithmic coordinates. Note that it decreases very rapidly for small
increases in its arguments. For example, when x increases from 4 to 5, Q(x) decreases by a factor of 100.
2
The term A T equals the energy expended by the transmitter in sending the bit; we
label this term Eb. We arrive at a concise expression for the probability the matched
filter receiver makes a bit-reception error.
(6.47)
Figure 6.16 shows how the receiver's error rate varies with the signal-to-noise ratio
291
Figure 6.16
The probability that the matched-filter receiver makes an error on any bit transmission is plotted against
the signal-to-noise ratio of the received signal. The upper curve shows the performance of the FSK signal
set, the lower (and therefore better) one the BPSK signal set.
Exercise 6.17.1
Derive the probability of error expression for the modulated BPSK signal set, and
show that its performance identically equals that of the baseband BPSK signal set.
6.18 Digital Communication System Properties
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Results from the Receiver Error module (Digital Communication in the Presence of
Noise (Page 289)) reveals several properties about digital communication systems.
• As the received signal becomes increasingly noisy, whether due to increased
distance from the transmitter (smaller α) or to increased noise in the channel
(larger N0), the probability the receiver makes an error approaches 1/2. In such
situations, the receiver performs only slightly better than the "receiver" that
ignores what was transmitted and merely guesses what bit was transmitted.
Consequently, it becomes almost impossible to communicate information when
digital channels become noisy.
• As the signal-to-noise ratio increases, performance gains smaller probability of
error pe can be easily obtained. At a signal-to-noise ratio of 12 dB, the probability
−8
the receiver makes an error equals 10 . In words, one out of one hundred
million bits will, on the average, be in error.
• Once the signal-to-noise ratio exceeds about 5 dB, the error probability decreases
dramatically. Adding 1 dB improvement in signal-to-noise ratio can result in a
factor of 10 smaller pe.
• Signal set choice can make a significant difference in performance. All BPSK signal
sets, baseband or modulated, yield the same performance for the same bit
292
energy. The BPSK signal set does perform much better than the FSK signal set
once the signal-to-noise ratio exceeds about 5 dB.
Exercise 6.18.1
Derive the expression for the probability of error that would result if the FSK signal set
were used.
The matched-filter receiver provides impressive performance once adequate signal-tonoise ratios occur. You might wonder whether another receiver might be better. The
answer is that the matched-filter receiver is optimal: No other receiver can provide a
smaller probability of error than the matched filter regardless of the SNR.
Furthermore, no signal set can provide better performance than the BPSK signal set,
where the signal representing a bit is the negative of the signal representing the other
bit. The reason for this result rests in the dependence of probability of error pe on the
difference between the noise-free integrator outputs: For a given Eb, no other signal
set provides a greater difference.
How small should the error probability be? Out of N transmitted bits, on the average
Npe bits will be received in error. Do note the phrase "on the average" here: Errors
occur randomly because of the noise introduced by the channel, and we can only
predict the probability of occurrence. Since bits are transmitted at a rate R, errors
occur at an average frequency of Rpe. Suppose the error probability is an impressively
−6
small number like 10 . Data on a computer network like Ethernet is transmitted at a
rate R = 100Mbps, which means that errors would occur roughly 100 per second. This
error rate is very high, requiring a much smaller pe to achieve a more acceptable
average occurrence rate for errors occurring. Because Ethernet is a wireline channel,
which means the channel noise is small and the attenuation low, obtaining very small
error probabilities is not difficult. We do have some tricks up our sleeves, however,
that can essentially reduce the error rate to zero without resorting to expending a
large amount of energy at the transmitter. We need to understand digital channels
(Digital Channels (Page 292) ) and Shannon's Noisy Channel Coding Theorem (Noisy
Channel Coding Theorem (Page 313) ).
6.19 Digital Channels
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Let's review how digital communication systems work within the Fundamental Model
of Communication (Figure 2.10: Fundamental model of communication). As shown in
Figure 6.17 (DigMC), the message is a single bit. The entire analog transmission/
reception system, which is discussed in Digital Communication (Digital
Communication (Page 281)), Signal Sets25, BPSK Signal Set26, Transmission
Bandwidth, Frequency Shift Keying (Frequency Shift Keying (Page 285) ), Digital
Communication Receivers (Digital Communication Receivers (Page 287)), Factors in
Receiver Error (Digital Communication in the Presence of Noise (Page 289)), Digital
Communication System Properties28, and Error Probability29, can be lumped into a
single system known as the digital channel.
293
Figure 6.17 DigMC
The steps in transmitting digital information are shown in the upper system, the Fundamental Model of
Communication. The symbolic-valued signal s (m) forms the message, and it is encoded into a bit
sequence b (n). The indices differ because more than one bit/symbol is usually required to represent the
message by a bitstream. Each bit is represented by an analog signal, transmitted through the (unfriendly)
channel, and received by a matched-filter receiver. From the received bitstreamb
the received symbolic-valued signals
is derived. The lower block diagram shows an equivalent system wherein the analog
portions are combined and modeled by a transition diagram, which shows how each
transmitted bit could be received. For example, transmitting a 0 results in the
reception of a 1 with probability pe (an error) or a 0 with probability 1 − pe (no error).
Digital channels are described by transition diagrams, which indicate the output
alphabet symbols that result for each possible transmitted symbol and the
probabilities of the various reception possibilities. The probabilities on transitions
coming from the same symbol must sum to one. For the matched-filter receiver and
the signal sets we have seen, the depicted transition diagram, known as a binary
symmetric channel, captures how transmitted bits are received. The probability of
error pe is the sole parameter of the digital channel, and it encapsulates signal set
choice, channel properties, and the matched-filter receiver. With this simple but
entirely accurate model, we can concentrate on how bits are received.
6.20 Entropy
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Communication theory has been formulated best for symbolic-valued signals. Claude
Shannon published in 1948 The Mathematical Theory of Communication, which
became the cornerstone of digital communication. He showed the power of
probabilistic models for symbolic-valued signals, which allowed him to quantify the
information present in a signal. In the simplest signal model, each symbol can occur at
index n with a probability Pr [ak], k = {1,...,K} . What this model says is that for each
signal value a K-sided coin is flipped (note that the coin need not be fair). For this
model to make sense, the probabilities must be numbers between zero and one and
must sum to one.
294
(6.48)
(6.49)
This coin-flipping model assumes that symbols occur without regard to what
preceding or succeeding symbols were, a false assumption for typed text. Despite this
probabilistic model's over-simplicity, the ideas we develop here also work when more
accurate, but still probabilistic, models are used. The key quantity that characterizes a
symbolic-valued signal is the entropy of its alphabet.
Because we use the base-2 logarithm, entropy has units of bits. For this Definition to
make sense, we must take special note of symbols having probability zero of
occurring. A zero-probability symbol never occurs; thus, we define 0log 20=0 so that
such symbols do not affect the entropy. The maximum value attainable by an
alphabet's entropy occurs when the symbols are equally likely
In this case, the entropy equals log2 K. The minimum value occurs when only one
symbol occurs; it has probability one of occurring and the rest have probability zero.
Exercise 6.20.1
Derive the maximum-entropy results, both the numeric aspect (entropy equals log 2K)
and the theoretical one (equally likely symbols maximize entropy). Derive the value of
the minimum entropy alphabet.
Example 6.1
A four-symbol alphabet has the following probabilities.
Note that these probabilities sum to one as they should. As
The entropy of this alphabet equals
295
(6.51)
6.21 Source Coding Theorem
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The significance of an alphabet's entropy rests in how we can represent it with a
sequence of bits. Bit sequences form the "coin of the realm" in digital
communications: they are the universal way of representing symbolic-valued signals.
We convert back and forth between symbols to bit-sequences with what is known as a
codebook: a table that associates symbols to bit sequences. In creating this table, we
must be able to assign a unique bit sequence to each symbol so that we can go
between symbol and bit sequences without error.
POINT OF INTEREST: You may be conjuring the notion of hiding information from
others when we use the name codebook for the symbol-to-bit-sequence table. There
is no relation to cryptology, which comprises mathematically provable methods of
securing information. The codebook terminology was developed during the
beginnings of information theory just after World War II.
As we shall explore in some detail elsewhere, digital communication (Section 6.13) is
the transmission of symbolic-valued signals from one place to another. When faced
with the problem, for example, of sending a file across the Internet, we must first
represent each character by a bit sequence. Because we want to send the file quickly,
we want to use as few bits as possible. However, we don't want to use so few bits that
the receiver cannot determine what each character was from the bit sequence. For
example, we could use one bit for every character: File transmission would be fast but
useless because the codebook creates errors. Shannon proved in his monumental
work what we call today the Source Coding Theorem. Let B (ak) denote the number
of bits used to represent the symbol ak. The average number of bits
required to represent the entire alphabet equals
The Source Coding Theorem states that the average number of bits needed to
accurately represent the alphabet need only to satisfy
(6.52)
296
Thus, the alphabet's entropy specifes to within one bit how many bits on the average
need to be used to send the alphabet. The smaller an alphabet's entropy, the fewer
bits required for digital transmission of files expressed in that alphabet.
Example 6.2
A four-symbol alphabet has the following probabilities.
and an entropy of 1.75 bits (Example 6.1). Let's see if we can find a codebook for this
four-letter alphabet that satisfies the Source Coding Theorem. The simplest code to try
is known as the simple binary code: convert the symbol's index into a binary number
and use the same number of bits for each symbol by including leading zeros where
necessary.
(6.53)
Whenever the number of symbols in the alphabet is a power of two (as in this case),
the average number of bits
equals log2K, which equals 2 in this case. Because the entropy equals 1.75bits, the
simple binary code indeed satisfes the Source Coding Theorem we are within one bit
of the entropy limit but you might wonder if you can do better. If we choose a
codebook with difering number of bits for the symbols, a smaller average number of
bits can indeed be obtained. The idea is to use shorter bit sequences for the symbols
that occur more often. One codebook like this is
(6.54)
Now
We can reach the entropy limit! The simple binary code is, in this case, less efcient than
the unequal-length code. Using the efcient code, we can transmit the symbolic-valued
signal having this alphabet 12.5% faster. Furthermore, we know that no more efcient
codebook can be found because of Shannon's Theorem.
297
6.22 Compression and the Huffman Code
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Shannon's Source Coding Theorem (6.52) has additional applications in data
compression. Here, we have a symbolic-valued signal source, like a computer file or
an image, that we want to represent with as few bits as possible. Compression
schemes that assign symbols to bit sequences are known as lossless if they obey the
Source Coding Theorem; they are lossy if they use fewer bits than the alphabet's
entropy. Using a lossy compression scheme means that you cannot recover a
symbolic-valued signal from its compressed version without incurring some error. You
might be wondering why anyone would want to intentionally create errors, but lossy
compression schemes are frequently used where the efficiency gained in representing
the signal outweighs the significance of the errors.
Shannon's Source Coding Theorem states that symbolic-valued signals require on the
average at least H(A) number of bits to represent each of its values, which are
symbols drawn from the alphabet A. In the module on the Source Coding Theorem
(Source Coding Theorem (Page 295)) we find that using a so-called fixed rate source
coder, one that produces a fixed number of bits/symbol, may not be the most efficient
way of encoding symbols into bits. What is not discussed there is a procedure for
designing an efficient source coder: one guaranteed to produce the fewest bits/
symbol on the average. That source coder is not unique, and one approach that does
achieve that limit is the Huffman source coding algorithm.
POINT OF INTEREST: In the early years of information theory, the race was on to be
the first to find a provably maximally efficient source coding algorithm. The race was
won by then MIT graduate student David Huffman in 1954, who worked on the
problem as a project in his information theory course. We're pretty sure he received
an "A."
• Create a vertical table for the symbols, the best ordering being in decreasing
order of probability.
• Form a binary tree to the right of the table. A binary tree always has two branches
at each node. Build the tree by merging the two lowest probability symbols at
each level, making the probability of the node equal to the sum of the merged
nodes' probabilities. If more than two nodes/symbols share the lowest probability
at a given level, pick any two; your choice won't affect
• At each node, label each of the emanating branches with a binary number. The
bit sequence obtained from passing from the tree's root to the symbol is its
Huffman code.
Example 6.3
The simple four-symbol alphabet used in the Entropy (Entropy (Page 293)) and Source
Coding (Source Coding Theorem (Page 295)) modules has a four-symbol alphabet with
the following probabilities,
298
and an entropy of 1.75 bits (Entropy (Page 293)). This alphabet has the Hufman coding
tree shown in Figure 6.18 (Hufman Coding Tree).
Figure 6.18 Huffman Coding Tree
We form a Huffman code for a four-letter alphabet having the indicated probabilities of occurrence. The
binary tree created by the algorithm extends to the right, with the root node (the one at which the tree
begins) defning the codewords. The bit sequence obtained by traversing the tree from the root to the
symbol defines that symbol's binary code.
The code thus obtained is not unique as we could have labeled the branches coming
out of each node differently. The average number of bits required to represent this
alphabet equals
1.75 bits, which is the Shannon entropy limit for this source alphabet. If we had the
symbolic-valued signal s (m)= {a2,a3,a1,a4,a1,a2,... }, our Hufman code would produce
the bitstream b (n) = 101100111010 ....
If the alphabet probabilities were different, clearly a different tree, and therefore
different code, could well result. Furthermore, we may not be able to achieve the
entropy limit. If our symbols had the probabilities
the average number of bits/symbol resulting from the Huffman coding algorithm
would equal 1.75 bits. However, the entropy limit is 1.68 bits. The Huffman code does
satisfy the Source Coding Theorem its average length is within one bit of the
alphabet's entropy but you might wonder if a better code existed. David Huffman
showed mathematically that no other code could achieve a shorter average code than
his. We can't do better.
Exercise 6.22.1
Derive the Huffman code for this second set of probabilities, and verify the claimed
average code length and alphabet entropy.
299
6.23 Subtlies of Coding
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
In the Huffman code, the bit sequences that represent individual symbols can have
differing lengths so the bitstream index m does not increase in lock step with the
symbol-valued signal's index n. To capture how often bits must be transmitted to keep
up with the source's production of symbols, we can only compute averages. If our
source code averages
bits/symbol and symbols are produced at a rate R, the average bit rate equals
, and this quantity determines the bit interval duration T.
Exercise 6.23.1
Calculate what the relation between T and the average bit rate
is.
A subtlety of source coding is whether we need "commas" in the bitstream. When we
use an unequal number of bits to represent symbols, how does the receiver
determine when symbols begin and end? If you created a source code that required a
separation marker in the bitstream between symbols, it would be very inefficient since
you are essentially requiring an extra symbol in the transmission stream.
NOTE: A good example of this need is the Morse Code: Between each letter, the
telegrapher needs to insert a pause to inform the receiver when letter boundaries
occur.
As shown in this example (Compression and the Huffman Code (Page 297)), no
commas are placed in the bitstream, but you can unambiguously decode the
sequence of symbols from the bitstream. Huffman showed that his (maximally
efficient) code had the prefix property: No code for a symbol began another symbol's
code. Once you have the prefix property, the bitstream is partially self-synchronizing:
Once the receiver knows where the bitstream starts, we can assign a unique and
correct symbol sequence to the bitstream.
Exercise 6.23.2
Sketch an argument that prefx coding, whether derived from a Hufman code or not,
will provide unique decoding when an unequal number of bits/symbol are used in the
code.
However, having a prefx code does not guarantee total synchronization: After hopping
into the middle of a bitstream, can we always find the correct symbol boundaries? The
self-synchronization issue does mitigate the use of efcient source coding algorithms.
Exercise 6.23.3
Show by example that a bitstream produced by a Hufman code is not necessarily selfsynchronizing. Are fxed-length codes self synchronizing?
300
Another issue is bit errors induced by the digital channel; if they occur (and they will),
synchronization can easily be lost even if the receiver started "in synch" with the
source. Despite the small probabilities of error ofered by good signal set design and
the matched filter, an infrequent error can devastate the ability to translate a
bitstream into a symbolic signal. We need ways of reducing reception errors without
demanding that pe be smaller.
Example 6.4
The first electrical communications system the telegraph was digital. When first
deployed in 1844, it communicated text over wireline connections using a binary code
the Morse code to represent individual letters. To send a message from one place to
another, telegraph operators would tap the message using a telegraph key to another
operator, who would relay the message on to the next operator, presumably getting
the message closer to its destination. In short, the telegraph relied on a network not
unlike the basics of modern computer networks. To say it presaged modern
communications would be an understatement. It was also far ahead of some needed
technologies, namely the Source Coding Theorem. The Morse code, shown in Figure
6.19, was not a prefix code. To separate codes for each letter, Morse code required
that a space a pause be inserted between each letter. In information theory, that
space counts as another code letter, which means that the Morse code encoded text
with a three-letter source code: dots, dashes and space. The resulting source code is
not within a bit of entropy, and is grossly inefficient (about 25%). Figure 6.19 shows a
Huffman code for English text, which as we know is efficient.
301
Figure 6.19 Morse and Huffman Code Table
Morse and Huffman Codes for American-Roman Alphabet. The % column indicates the average
probability (expressed in percent) of the letter occurring in English. The entropy H (A) of the this source is
4.14 bits. The average Morse codeword length is 2.5 symbols. Adding one more symbol for the letter
separator and converting to bits yields an average codeword length of 5.56 bits. The average Huffman
codeword length is 4.35 bits.
302
6.24 Channel Coding
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
We can, to some extent, correct errors made by the receiver with only the error-filled
bit stream emerging from the digital channel available to us. The idea is for the
transmitter to send not only the symbol-derived bits emerging from the source coder
but also additional bits derived from the coder's bit stream. These additional bits, the
error correcting bits, help the receiver determine if an error has occurred in the data
bits (the important bits) or in the error-correction bits. Instead of the communication
model (Figure 6.17: DigMC) shown previously, the transmitter inserts a channel coder
before analog modulation, and the receiver the corresponding channel decoder
(Figure 6.20). This block diagram shown there forms the Fundamental Model of
Digital Communication.
Figure 6.20
To correct errors that occur in the digital channel, a channel coder and decoder are
added to the communication system. Properly designed channel coding can greatly
reduce the probability (from the uncoded value of pe) that a data bit b (n) is received
incorrectly even when the probability of c (l) be received in error remains pe or
becomes larger. This system forms the Fundamental Model of Digital Communication.
Shannon's Noisy Channel Coding Theorem (Noisy Channel Coding Theorem (Page 313)
) says that if the data aren't transmitted too quickly, that error correction codes exist
that can correct all the bit errors introduced by the channel. Unfortunately, Shannon
did not demonstrate an error correcting code that would achieve this remarkable feat;
in fact, no one has found such a code. Shannon's result proves it exists; seems like
there is always more work to do. In any case, that should not prevent us from studying
commonly used error correcting codes that not only find their way into all digital
communication systems, but also into CDs and bar codes used on merchandise.
6.25 Repetition Codes
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Perhaps the simplest error correcting code is the repetition code.
303
Figure 6.21 Repetition Code
The upper portion depicts the result of directly modulating the bit stream b (n) into a transmitted signal x (t)
using a baseband BPSK signal set. R is the datarate produced by the source coder. If that bit stream
passes through a (3,1) channel coder to yield the bit stream c (l), the resulting transmitted signal requires
a bit interval T three times smaller than the uncoded version. This reduction in the bit interval means that
the transmitted energy/bit decreases by a factor of three, which results in an increased error probability in
the receiver.
Here, the transmitter sends the data bit several times, an odd number of times in fact.
Because the error probability pe is always less than
we know that more of the bits should be correct rather than in error. Simple majority
voting of the received bits (hence the reason for the odd number) determines the
transmitted bit more accurately than sending it alone. For example, let's consider the
three-fold repetition code: for every bit b (n) emerging from the source coder, the
channel coder produces three. Thus, the bit stream emerging from the channel coder
c (l) has a data rate three times higher than that of the original bit stream b (n). The
coding table illustrates when errors can be corrected and when they can't by the
majority-vote decoder.
Code
Probability
Bit
000
0
001
0
010
0
011
1
100
0
101
1
110
1
111
1
Table 6.1 Coding Table
304
In this example, the transmitter encodes 0 as 000. The channel creates an error (changing a 0 into a 1)
that with probability pe. The first column lists all possible received datawords and the second the
probability of each dataword being received. The last column shows the results of the majority-vote
decoder. When the decoder produces 0, it successfully corrected the errors introduced by
the channel (if there were any; the top row corresponds to the case in which no errors
occurred). The error probability of the decoders is the sum of the probabilities when the
decoder produces 1.
Thus, if one bit of the three bits is received in error, the receiver can correct the error;
if more than one error occurs, the channel decoder announces the bit is 1 instead of
transmitted value of 0. Using this repetition code, the probability of
This probability of a decoding error is always less than pe, the uncoded value, so long as
Exercise 6.25.1
Demonstrate mathematically that this claim is indeed true. Is
6.26 Block Channel Coding
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Because of the higher datarate imposed by the channel coder, the probability of bit
error occurring in the digital channel increases relative to the value obtained when no
channel coding is used. The bit interval duration must be reduced by
in comparison to the no-channel-coding situation, which means the energy per bit Eb
goes down by the same amount. The bit interval must decrease by a factor of three if
the transmitter is to keep up with the data stream, as illustrated here (Figure 6.21:
Repetition Code).
POINT OF INTEREST: It is unlikely that the transmitter's power could be increased to
compensate.
Such is the sometimes-unfriendly nature of the real world.
Because of this reduction, the error probability pe of the digital channel goes up. The
question thus becomes does channel coding really help: Is the effective error
probability lower with channel coding even though the error probability for each
transmitted bit is larger? The answer is no: Using a repetition code for channel coding
cannot ultimately reduce the probability that a data bit is received in error. The
ultimate reason is the repetition code's inefficiency: transmitting one data bit for every
three transmitted is too inefficient for the amount of error correction provided.
Exercise 6.26.1
305
Using MATLAB, calculate the probability a bit is received incorrectly with a three-fold
repetition code. Show that when the energy per bit Eb is reduced by 1/3 that this
probability is larger than the no-coding probability of error.
Repetition Codes (Page 302) represents a special case of what is known as block
channel coding. For every K bits that enter the block channel coder, it inserts an
additional N - K error-correction bits to produce a block of N bits for transmission. We
use the notation (N,K) to represent a given block code's parameters. In the three-fold
Repetition Codes (Page 302), K = 1 and N = 3. A block code's coding efficiency E equals
the ratio
, and quantifies the overhead introduced by channel coding. The rate at which bits
must be transmitted again changes: So-called data bits b (n) emerge from the source
coder at an average rate
and exit the channel at a rate
higher. We represent the fact that the bits sent through the digital channel operate at
a different rate by using the index l for the channel-coded bit stream c (l). Note that
the blocking (framing) imposed by the channel coder does not correspond to symbol
boundaries in the bit stream b (n), especially when we employ variable-length source
codes.
Does any error-correcting code reduce communication errors when real-world
constraints are taken into account? The answer now is yes. To understand channel
coding, we need to develop first a general framework for channel coding, and discover
what it takes for a code to be maximally efficient: Correct as many errors as possible
using the fewest error correction bits as possible (making the efficiency
as large as possible).
6.27 Error-Correcting Codes: Hamming Distance
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
So-called linear codes create error-correction bits by combining the data bits linearly.
The phrase "linear combination" means here single-bit binary arithmetic.
For example, let's consider the specific (3, 1) error correction code described by the
following coding table and, more concisely, by the succeeding matrix expression.
c (1) = b (1)
306
c (2) = b (1)
c(3) = b (1)
or
c = Gb
where
The length-K (in this simple example K =1) block of data bits is represented by the
vector b, and the length-N output block of the channel coder, known as a codeword,
by c. The generator matrixG defines all block-oriented linear channel coders.
As we consider other block codes, the simple idea of the decoder taking a majority
vote of the received bits won't generalize easily. We need a broader view that takes
into account the distance between codewords. A length-N codeword means that the
N
receiver must decide among the 2 possible datawords to select which of the
K
2 codewords was actually transmitted. As shown in Figure 6.22, we can think of the
datawords geometrically. We define the Hamming distance between binary
datawords c1 and c2, denoted by d (c1,c2) to be the minimum number of bits that must
be "fipped" to go from one word to the other. For example, the distance between
codewords is 3 bits. In our table of binary arithmetic, we see that adding a 1
corresponds to fipping a bit. Furthermore, subtraction and addition are equivalent.
We can express the Hamming distance as
d (c1,c2) = sum(c1 ⊕ c2)
(6.55)
Exercise 6.27.1
Show that adding the error vector col [1,0,...,01 to a codeword flips the codeword's
leading bit and leaves the rest unaffected.
The probability of one bit being flipped anywhere in a codeword is
The number of errors the channel introduces equals the number of ones in e; the
probability of any particular error vector decreases with the number of errors.
Figure 6.22
307
In a (3,1) repetition code, only 2 of the possible 8 three-bit data blocks are codewords. We can represent
these bit patterns geometrically with the axes being bit positions in the data block. In the left plot, the filled
circles represent the codewords [0 0 0] and [1 1 1], the only possible codewords. The unflled ones
correspond to the transmission. The center plot shows that the distance between codewords is 3.
Because distance corresponds to flipping a bit, calculating the Hamming distance geometrically means
following the axes rather than going "as the crow flies". The right plot shows the datawords that result
when one error occurs as the codeword goes through the channel. The three datawords are unit distance
from the original codeword. Note that the received dataword groups do not overlap, which means the code
can correct all single-bit errors.
To perform decoding when errors occur, we want to find the codeword (one of the
filled circles in Figure 6.22) that has the highest probability of occurring: the one
closest to the one received. Note that if a dataword lies a distance of 1 from two
codewords, it is impossible to determine which codeword was actu ally sent. This
criterion means that if any two codewords are two bits apart, then the code cannot
correct the channel-induced error. Thus, to have a code that can correct all singlebit errors, codewords must have a minimum separation of three. Our repetition
code has this property.
Introducing code bits increases the probability that any bit arrives in error (because bit
interval durations decrease). However, using a well-designed error-correcting code
corrects bit reception errors. Do we win or lose by using an error-correcting code? The
answer is that we can win if the code is well-designed. The (3,1) repetition code
demonstrates that we can lose (Block Channel Coding (Page 304)). To develop good
channel coding, we need to develop first a general framework for channel codes and
discover what it takes for a code to be maximally efficient: Correct as many errors as
possible using the fewest error correction bits as possible (making the K efficiency
as large as possible.) We also need a systematic way of finding the codeword closest
to any received dataword. A much better code than our (3,1) repetition code is the
following (7,4) code.
where the generator matrix is
308
4
7
In this (7,4) code, 2 = 16 of the 2 = 128 possible blocks at the channel decoder
correspond to error-free transmission and reception.
Error correction amounts to searching for the codeword c closest to the received block
c in terms of the Hamming distance between the two. The error correction capability
of a channel code is limited by how close together any two error-free blocks are. Bad
codes would produce blocks close together, which would result in ambiguity when
assigning a block of data bits to a received block. The quantity to examine, therefore,
in designing code error correction codes is the minimum distance between
codewords.
(6.56)
To have a channel code that can correct all single-bit errors, dmin ≥ 3.
Exercise 6.27.2
Suppose we want a channel code to have an error-correction capability of n bits. What
must the minimum Hamming distance between codewords dmin be?
How do we calculate the minimum distance between codewords? Because we have 2
K−1
K
K–1
codewords, the number of possible unique pairs equals 2 (2 ) , which can be a
large number. Recall that our channel coding procedure is linear, with c = Gb.
Therefore ci ⊕ cj = G (bi ⊕ bj). Because bi ⊕ bj always yields another block of data bits,
we find that the difference between any two codewords is another codeword! Thus, to
find dmin we need only compute the number of ones that comprise all non-zero
codewords. Finding these codewords is easy once we examine the coder's generator
matrix. Note that the columns of G are codewords (why is this?), and that all
codewords can be found by all possible pairwise sums of the columns. To find dmin ,
we need only count the number of bits in each column and sums of columns. For our
example (7, 4), G's first column has three ones, the next one four, and the last two
three. Considering sums of column pairs next, note that because the upper portion of
G is an identity matrix, the corresponding upper portion of all column sums must have
exactly two bits. Because the bottom portion of each column differs from the other
columns in at least one place, the bottom portion of a sum of columns must have at
least one bit. Triple sums will have at least three bits because the upper portion of G is
an identity matrix. Thus, no sum of columns has fewer than three bits, which means
that dmin =3, and we have a channel coder that can correct all occurrences of one
error within a received 7-bit block.
6.28 Error-Correcting Codes: Channel Decoding
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Because the idea of channel coding has merit (so long as the code is efficient), let's
develop a systematic procedure for performing channel decoding. One way of
checking for errors is to try recreating the error correction bits from the data portion
309
of the received block c. Using matrix notation, we make this calculation by multiplying
the received block c by the matrix H known as the parity check matrix. It is formed
from the generator matrix G by taking the bottom, error-correction portion of G and
attaching to it an identity matrix. For our (7,4) code,
(6.57)
The parity check matrix thus has size (N - K) × N, and the result of multiplying this
matrix with a received word is a length- (N - K) binary vector. If no digital channel
errors occur we receive a codeword so that
T
For example, the first column of G, (1, 0, 0, 0, 1, 0, 1) , is a codeword. Simple
calculations show that multiplying this vector by H results in a length-(N - K) zerovalued vector.
Exercise 6.28.1
Show that Hc =0 for all the columns of G. In other words, show that HG =0 an (N - K) ×
K matrix of zeroes. Does this property guarantee that all codewords also satisfy Hc=0?
When the received bits
do not form a codeword,
does not equal zero, indicating the presence of one or more errors induced by the
digital channel. Because the presence of an error can be mathematically written as
with e a vector of binary values having a 1 in those positions where a bit error
occurred.
Exercise 6.28.2
T
Show that adding the error vector (1, 0,..., 0) to a codeword fips the codeword's
leading bit and leaves the rest unaffected.
Consequently,
Because the result of the product is a length-(N − K) vector of binary values, we can
N−K
have 2
− 1 non-zero values that correspond to non-zero error patterns e. To
perform our channel decoding,
1. compute (conceptually at least)
2. if this result is zero, no detectable or correctable error occurred;
3. if non-zero, consult a table of length-(N − K) binary vectors to associate them with
the minimal error pattern that could have resulted in the non-zero result; then
310
4. add the error vector thus obtained to the received vector
to correct the error (because c ⊕ e ⊕ e =c).
5. Select the data bits from the corrected word to produce the received bit sequence
The phrase minimal in the third item raises the point that a double (or triple or
quadruple ...) error occurring during the transmission/reception of one codeword can
create the same received word as a single-bit error or no error in another codeword.
T
T
For example, (1, 0, 0, 0, 1, 0, 1) and (0, 1, 0, 0, 1, 1, 1) are both codewords in the
example (7,4) code. The second results when the first one experiences three bit errors
(first, second, and sixth bits). Such an error pattern cannot be detected by our coding
strategy, but such multiple error patterns are very unlikely to occur. Our receiver uses
the principle of maximum probability: An error-free transmission is much more likely
than one with three errors if the bit-error probability pe is small enough.
Exercise 6.28.3
How small must pe be so that a single-bit error is more likely to occur than a triple-bit
error?
6.29 Error-Correcting Codes: Hamming Codes
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
N-K
For the (7,4) example, we have 2
− 1=7 error patterns that can be corrected. We
start with single-bit error patterns, and multiply them by the parity check matrix. If we
obtain unique answers, we are done; if two or more error patterns yield the same
result, we can try double-bit error patterns. In our case, single-bit error patterns give a
unique result.
311
e
He
1000000
101
0100000
111
0010000
110
0001000
011
0000100
100
0000010
010
0000001
001
Table 6.2 Parity Check Matrix
Table 6.3
This corresponds to our decoding table: We associate the parity check matrix
multiplication result with the error pattern and add this to the received word. If more
than one error occurs (unlikely though it may be), this "error correction" strategy
usually makes the error worse in the sense that more bits are changed from what was
transmitted.
As with the repetition code, we must question whether our (7,4) code's error
correction capability compensates for the increased error probability due to the
necessitated reduction in bit energy. Figure 6.23 (Probability of error occurring) shows
that if the signal-to-noise ratio is large enough channel coding yields a smaller error
probability. Because the bit stream emerging from the source decoder is segmented
into four-bit blocks, the fair way of comparing coded and uncoded transmission is to
compute the probability of block error: the probability that any bit in a block remains
in error despite error correction and regardless of whether the error occurs in the
data or in coding buts. Clearly, our (7,4) channel code does yield smaller error rates,
and is worth the additional systems required to make it work.
312
Figure 6.23 Probability of error occurring
4
4
The probability of an error occurring in transmitted K =4 data bits equals 1 − (1 − pe) as (1 − pe) equals
the probability that the four bits are received without error. The upper curve displays how this probability of
an error anywhere in the four-bit block varies with the signal-to-noise ratio. When a (7,4) single-bit error
correcting code is used, the transmitter reduced the energy it expends during a single-bit transmission by
4/7, appending three extra bits for error correction. Now the probability of any bit in the seven-bit block
being in error after error correction equals
where p'e is the probability of a bit error occurring in the channel when channel coding
occurs. Here
equals the probability of exactly on in seven bits emerging from the channel in error;
The channel decoder corrects this type of error, and all data bits in the block are
received correctly.
Note that our (7,4) code has the length and number of data bits that perfectly fts
correcting single bit errors. This pleasant property arises because the number of error
patterns that can be corrected, 2
N−K
N−K
− 1, equals the codeword length N. Codes that
have 2
− 1= N are known as Hamming codes, and the following table (Table 6.3:
Hamming Codes) provides the parameters of these codes. Hamming codes are the
simplest single-bit error correction codes, and the generator/parity check matrix
formalism for channel coding and decoding works for them.
313
N
K
E (efficiency)
3
1
0.33
7
4
0.57
15
11
0.73
31
26
0.84
63
57
0.90
127
120
0.94
Table 6.3 Hamming Codes
Unfortunately, for such large blocks, the probability of multiple-bit errors can exceed
the number of single-bit errors unless the channel single-bit error probability Pe is very
small. Consequently, we need to enhance the code's error correcting capability by
adding double as well as single-bit error correction.
Exercise 6.29.1
What must the relation between N and K be for a code to correct all single-and doublebit errors with a "perfect fit"?
6.30 Noisy Channel Coding Theorem
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
As the block length becomes larger, more error correction will be needed. Do codes
exist that can correct all errors? Perhaps the crowning achievement of Claude
Shannon's creation of information theory answers this question. His result comes in
two complementary forms: the Noisy Channel Coding Theorem and its converse.
6.30.1 Noisy Channel Coding Theorem
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Let E denote the efficiency of an error-correcting code: the ratio of the number of data
bits to the total number of bits used to represent them. If the efficiency is less than
the capacity of the digital channel, an error-correcting code exists that has the
property that as the length of the code increases, the probability of an error occurring
in the decoded block approaches zero.
314
(6.58)
6.30.2 Converse to the Noisy Channel Coding Theorem
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
If E > C, the probability of an error in a decoded block must approach one regardless
of the code that might be chosen.
(6.59)
These results mean that it is possible to transmit digital information over a noisy
channel (one that introduces errors) and receive the information without error if the
code is sufficiently inefficient compared to the channel's characteristics. Generally, a
channel's capacity changes with the signal-to-noise ratio: As one increases or
decreases, so does the other. The capacity measures the overall error characteristics
of a channel the smaller the capacity the more frequently errors occur and an overly
efficient error-correcting code will not build in enough error correction capability to
counteract channel errors.
This result astounded communication engineers when Shannon published it in 1948.
Analog communication always yields a noisy version of the transmitted signal; in
digital communication, error correction can be powerful enough to correct all errors
as the block length increases. The key for this capability to exist is that the code's
efficiency be less than the channel's capacity. For a binary symmetric channel, the
capacity is given by
(6.60)
Figure 6.24 (capacity of a channel) shows how capacity varies with error probability.
For example, our (7,4) Hamming code has an efficiency of 0.57, and codes having the
same efciency but longer block sizes can be used on additive noise channels where
the signal-to-noise ratio exceeds 0dB.
315
Figure 6.24 Capacity of a channel
The capacity per transmission through a binary symmetric channel is plotted as a function of the digital
channel's error probability (upper) and as a function of the signal-to-noise ratio for a BPSK signal set
(lower).
6.31 Capacity of a Channel
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
In addition to the Noisy Channel Coding Theorem and its converse (Section 6.30),
Shannon also derived the capacity for a bandlimited (to W Hz) additive white noise
channel. For this case, the signal set is unrestricted, even to the point that more than
one bit can be transmitted each "bit interval." Instead of constraining channel code
efficiency, the revised Noisy Channel Coding Theorem states that some errorcorrecting code exists such that as the block length increases, error-free transmission
is possible if the source coder's datarat,
is less than capacity.
(6.61)
This result sets the maximum datarate of the source coder's output that can be
transmitted through the bandlimited channel with no error. Shannon's proof of his
theorem was very clever, and did not indicate what this code might be; it has never
been found. Codes such as the Hamming code work quite well in practice to keep
error rates low, but they remain greater than zero. Until the "magic" code is found,
more important in communication system design is the converse. It states that if your
data rate exceeds capacity, errors will overwhelm you no matter what channel coding
you use. For this reason, capacity calculations are made to understand the
fundamental limits on transmission rates.
Exercise 6.31.1
316
The first Definition of capacity applies only for binary symmetric channels, and
represents the number of bits/transmission. The second result states capacity more
generally, having units of bits/second. How would you convert the first definition’s
result into units of bits/second?
Example 6.5
The telephone channel has a bandwidth of 3 kHz and a signal-to-noise ratio exceeding
30 dB (at least they promise this much). The maximum data rate a modem can
produce for this wireline channel and hope that errors will not become rampant is the
capacity.
(6.62)
Thus, the so-called 33 kbps modems operate right at the capacity limit.
Note that the data rate allowed by the capacity can exceed the bandwidth when the
signal-to-noise ratio exceeds 0 dB. Our results for BPSK and FSK indicated the
bandwidth they require exceeds
What kind of signal sets might be used to achieve capacity? Modem signal sets send
more than one bit/transmission using a number, one of the most popular of which is
multi-level signaling. Here, we can transmit several bits during one transmission
interval by representing bit by some signal's amplitude. For example, two bits can be
sent with a signal set comprised of a sinusoid with amplitudes of
6.32 Comparison of Analog and Digital Communication
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Analog communication systems, amplitude modulation (AM) radio being a typifying
example, can inexpensively communicate a bandlimited analog signal from one
location to another (point-to-point communication) or from one point to many
(broadcast). Although it is not shown here, the coherent receiver (Figure 6.6) provides
the largest possible signal-to-noise ratio for the demodulated message. An analysis
(Section 6.12) of this receiver thus indicates that some residual error will always be
present in an analog system's output.
Although analog systems are less expensive in many cases than digital ones for the
same application, digital systems offer much more efficiency, better performance, and
much greater flexibility.
• Efficiency: The Source Coding Theorem allows quantifcation of just how complex
a given message source is and allows us to exploit that complexity by source
coding (compression). In analog communication, the only parameters of interest
317
are message bandwidth and amplitude. We cannot exploit signal structure to
achieve a more efcient communication system.
• Performance: Because of the Noisy Channel Coding Theorem, we have a specifc
criterion by which to formulate error-correcting codes that can bring us as close
to error-free transmission as we might want. Even though we may send
information by way of a noisy channel, digital schemes are capable of error-free
transmission while analog ones cannot overcome channel disturbances; see this
problem (Information Communication Problems (Page 325)) for a comparison.
• Flexibility: Digital communication systems can transmit real-valued discrete-time
signals, which could be analog ones obtained by analog-to-digital conversion, and
symbolic-valued ones (computer data, for example). Any signal that can be
transmitted by analog means can be sent by digital means, with the only issue
being the number of bits used in A/D conversion (how accurately do we need to
represent signal amplitude). Images can be sent by analog means (commercial
television), but better communication performance occurs when we use digital
systems (HDTV). In addition to digital communication's ability to transmit a wider
variety of signals than analog systems, point-to-point digital systems can be
organized into global (and beyond as well) systems that provide efficient and
flexible information transmission. Computer networks, explored in the next
section, are what we call such systems today. Even analog-based networks, such
as the telephone system, employ modern computer networking ideas rather than
the purely analog systems of the past.
Consequently, with the increased speed of digital computers, the development of
increasingly efficient algorithms, and the ability to interconnect computers to form a
communications infrastructure, digital commu nication is now the best choice for
many situations.
6.33 Communication Networks
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Communication networks elaborate the Fundamental Model of Communications
(Figure 1.4: Fundamental model of communication). The model shown in Figure
6.25 describes point-to-point communications well, wherein the link between
transmitter and receiver is straightforward, and they have the channel to themselves.
One modern example of this communications mode is the modem that connects a
personal computer with an information server via a telephone line. The key aspect,
some would say faw, of this model is that the channel is dedicated: Only one
communications link through the channel is allowed for all time. Regardless whether
we have a wireline or wireless channel, communication bandwidth is precious, and if it
could be shared without significant degradation in communications performance
(measured by signal-to-noise ratio for analog signal transmission and by bit-error
probability for digital transmission) so much the better.
318
Figure 6.25
The prototypical communications network whether it be the postal service, cellular telephone, or the
Internet consists of nodes interconnected by links. Messages formed by the source are transmitted within
the network by dynamic routing. Two routes are shown. The longer one would be used if the direct link
were disabled or congested.
The idea of a network first emerged with perhaps the oldest form of organized
communication: the postal service. Most communication networks, even modern
ones, share many of its aspects.
• A user writes a letter, serving in the communications context as the message
source.
• This message is sent to the network by delivery to one of the network's public
entry points. Entry points in the postal case are mailboxes, post offices, or your
friendly mailman or mailwoman picking up the letter.
• The communications network delivers the message in the most efficient (timely)
way possible, trying not to corrupt the message while doing so.
• The message arrives at one of the network's exit points, and is delivered to the
recipient (what we have termed the message sink).
Exercise 6.33.1
Develop the network model for the telephone system, making it as analogous as
possible with the postal service-communications network metaphor. What is most
interesting about the network system is the ambivalence of the message source and
sink about how the communications link is made. What they do care about is message
integrity and communications efficiency. Furthermore, today's networks use
heterogeneous links. Communication paths that form the Internet use wireline, optical
fiber, and satellite communication links.
The first electrical communications network was the telegraph. Here the network
consisted of telegraph operators who transmitted the message efficiently using Morse
code and routed the message so that it took the shortest possible path to its
destination while taking into account internal network failures (downed lines, drunken
operators). From today's perspective, the fact that this nineteenth century system
handled digital communications is astounding. Morse code, which assigned a
sequence of dots and dashes to each letter of the alphabet, served as the source
coding algorithm. The signal set consisted of a short and a long pulse. Rather than a
matched filter, the receiver was the operator's ear, and he wrote the message
(translating from received bits to symbols).
NOTE: Because of the need for a comma between dot-dash sequences to define letter
(symbol) boundaries, the average number of bits/symbol, as described in Subtleties of
319
Coding (Subtlies of Coding (Page 299)), exceeded the Source Coding Theorem's
upper bound.
Internally, communication networks do have point-to-point communication links
between network nodes well described by the Fundamental Model of
Communications. However, many messages share the communications channel
between nodes using what we call time-domain multiplexing: Rather than the
continuous communications mode implied in the Model as presented, message
sequences are sent, sharing in time the channel's capacity. At a grander viewpoint, the
network must route messages decide what nodes and links to use based on
destination information the address that is usually separate from the message
information. Routing in networks is necessarily dynamic: The complete route taken by
messages is formed as the network handles the message, with nodes relaying the
message having some notion of the best possible path at the time of transmission.
Note that no omnipotent router views the network as a whole and pre-determines
every message's route. Certainly in the case of the postal system dynamic routing
occurs, and can consider issues like inoperative and overly busy links. In the telephone
system, routing takes place when you place the call; the route is fixed once the phone
starts ringing. Modern communication networks strive to achieve the most efficient
(timely) and most reliable information delivery system possible.
6.34 Message Routing
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Focusing on electrical networks, most analog ones make inefficient use of
communication links because truly dynamic routing is difficult, if not impossible, to
obtain. In radio networks, such as commercial television, each station has a dedicated
portion of the electromagnetic spectrum, and this spectrum cannot be shared with
other stations or used in any other than the regulated way. The telephone network is
more dynamic, but once it establishes a call the path through the network is fixed. The
users of that path control its use, and may not make efficient use of it (long pauses
while one person thinks, for example). Telephone network customers would be quite
upset if the telephone company momentarily disconnected the path so that someone
else could use it. This kind of connection through a network fixed for the duration of
the communication session is known as a circuit-switched connection.
During the 1960s, it was becoming clear that not only was digital communication
technically superior, but also that the wide variety of communication modes computer
login, file transfer, and electronic mail needed a different approach than point-topoint. The notion of computer networks was born then, and what was then called the
ARPANET, now called the Internet, was born. Computer networks elaborate the basic
network model by subdividing messages into smaller chunks called packets (Figure
6.26). The rationale for the network enforcing smaller transmissions was that large file
transfers would consume network resources all along the route, and, because of the
long transmission time, a communication failure might require retransmission of the
entire file. By creating packets, each of which has its own address and is routed
independently of others, the network can better manage congestion. The analogy is
320
that the postal service, rather than sending a long letter in the envelope you provide,
opens the envelope, places each page in a separate envelope, and using the address
on your envelope, addresses each page's envelope accordingly, and mails them
separately. The network does need to make sure packet sequence (page numbering)
is maintained, and the network exit point must reassemble the original message
accordingly.
Figure 6.26
Long messages, such as files, are broken into separate packets, then transmitted over computer
networks. A packet, like a letter, contains the destination address, the return address (trans mitter
address), and the data. The data includes the message part and a sequence number identifying its order
in the transmitted message.
Communications networks are now categorized according to whether they use
packets or not. A system like the telephone network is said to be circuit switched:
The network establishes a fixed route that lasts the entire duration of the message.
Circuit switching has the advantage that once the route is determined, the users can
use the capacity provided them however they like. Its main disadvantage is that the
users may not use their capacity efficiently, clogging network links and nodes along
the way. Packet-switched networks continuously monitor network utilization and
route messages accordingly. Thus, messages can, on the average, be delivered
efficiently, but the network cannot guarantee a specific amount of capacity to the
users.
6.35 Network architectures and interconnection
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The network structure its architecture (Figure 6.25) typifes what are known as wide
area networks (WANs). The nodes, and users for that matter, are spread
geographically over long distances. "Long" has no precise defnition, and is intended to
suggest that the communication links vary widely. The Internet is certainly the largest
WAN, spanning the entire earth and beyond. Local area networks, LANs, employ a
single communication link and special routing. Perhaps the best known LAN is
Ethernet. LANs connect to other LANs and to wide area networks through special
nodes known as gateways (Figure 6.27). In the Internet, a computer's address
consists of a four byte sequence, which is known as its IP address (Internet Protocol
address). An example address is 128.42.4.32: each byte is separated by a period. The
first two bytes specify the computer's domain (here Rice University). Computers are
321
also addressed by a more human-readable form: a sequence of alphabetic
abbreviations representing institution, type of institution, and computer name. A given
computer has both names (128.42.4.32 is the same as soma.rice.edu). Data
transmission on the Internet requires the numerical form. So-called name servers
translate between alphabetic and numerical forms, and the transmitting computer
requests this translation before the message is sent to the network.
Figure 6.27
The gateway serves as an interface between local area networks and the Internet. The two shown here
translate between LAN and WAN protocols; one of these also interfaces between two LANs, presumably
because together the two LANs would be geographically too dispersed.
6.36 Ethernet
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Figure 6.28
The Ethernet architecture consists of a single coaxial cable terminated at either end by a resistor having a
value equal to the cable's characteristic impedance. Computers attach to the Ethernet through an
interface known as a transceiver because it sends as well as receives bit streams represented as analog
voltages.
Ethernet uses as its communication medium a single length of coaxial cable (Figure
6.28). This cable serves as the "ether", through which all digital data travel. Electrically,
computers interface to the coaxial cable (Figure 6.28) through a device known as a
transceiver. This device is capable of monitoring the voltage appearing between the
core conductor and the shield as well as applying a voltage to it. Conceptually it
consists of two op-amps, one applying a voltage corresponding to a bit stream
(transmitting data) and another serving as an amplifier of Ethernet voltage signals
(receiving data). The signal set for Ethernet resembles that shown in BPSK Signal Sets,
322
with one signal the negative of the other. Computers are attached in parallel, resulting
in the circuit model for Ethernet shown in Figure 6.29.
Exercise 6.36.1
From the viewpoint of a transceiver's sending op-amp, what is the load it sees and
what is the transfer function between this output voltage and some other transceiver's
receiving circuit? Why should the output resistor Rout be large?
Figure 6.29
The top circuit expresses a simplified circuit model for a transceiver. The output resistance Rout must be
much larger than Z0 so that the sum of the various transmitter voltages add to create the Ethernet
conductor-to-shield voltage that serves as the received signal r (t) for all transceivers. In this case, the
equivalent circuit shown in the bottom circuit applies.
No one computer has more authority than any other to control when and how
messages are sent. Without scheduling authority, you might well wonder how one
computer sends to another without the (large) interference that the other computers
would produce if they transmitted at the same time. The innovation of Ethernet is that
computers schedule themselves by a random-access method. This method relies on
the fact that all packets transmitted over the coaxial cable can be received by all
transceivers, regardless of which computer might actually be the intended recipient. In
communications terminology, Ethernet directly supports broadcast. Each computer
goes through the following steps to send a packet.
1. The computer senses the voltage across the cable to determine if some other
computer is transmitting.
2. If another computer is transmitting, wait until the transmissions finish and go
back to the first step. If the cable has no transmissions, begin transmitting the
packet.
3. If the receiver portion of the transceiver determines that no other computer is
also sending a packet, continue transmitting the packet until completion.
4. On the other hand, if the receiver senses interference from another computer's
transmissions, immediately cease transmission, waiting a random amount of time
323
to attempt the transmission again (go to step 1) until only one computer
transmits and the others defer. The condition wherein two (or more) computers'
transmissions interfere with others is known as a collision.
The reason two computers waiting to transmit may not sense the other's transmission
immediately arises because of the finite propagation speed of voltage signals through
the coaxial cable. The longest time any computer must wait to determine if its
transmissions do not encounter interference is
where L is the coaxial cable's length. The maximum-length-specification for Ethernet
is 1 km. Assuming a propagation speed of 2/3 the speed of light, this time interval is
more than 10 µs. As analyzed in Problem 22 (Information Communication Problems
(Page 325)), the number of these time intervals required to resolve the collision is, on
the average, less than two!
Exercise 6.36.2
Why does the factor of two enter into this equation? (Consider the worst-case
situation of two transmitting computers located at the Ethernet's ends.)
Thus, despite not having separate communication paths among the computers to
coordinate their transmissions, the Ethernet random access protocol allows
computers to communicate without only a slight degradation in efficiency, as
measured by the time taken to resolve collisions relative to the time the Ethernet is
used to transmit information.
A subtle consideration in Ethernet is the minimum packet size Pmin. The time required
to transmit such packets equals
where C is the Ethernet's capacity in bps. Ethernet now comes in two different types,
C each with individual specifications, the most distinguishing of which is capacity: 10
Mbps and 100 Mbps. If the minimum transmission time is such that the beginning of
the packet has not propagated the full length of the Ethernet before the end-oftransmission, it is possible that two computers will begin transmission at the same
time and, by the time their transmissions cease, the other's packet will not have
propagated to the other. In this case, computers in-between the two will sense a
collision, which renders both computer's transmissions senseless to them, without the
two transmitting computers knowing a collision has occurred at all! For Ethernet to
succeed, we must have the minimum packet transmission time exceed twice the
voltage propagation time:
or
(6.63)
Thus, for the 10 Mbps Ethernet having a 1 km maximum length specifcation, the
minimum packet size is 200 bits.
324
Exercise 6.36.3
The 100 Mbps Ethernet was designed more recently than the 10 Mbps alternative. To
maintain the same minimum packet size as the earlier, slower version, what should its
length specification be? Why should the minimum packet size remain the same?
6.37 Communication Protocols
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The complexity of information transmission in a computer network reliable
transmission of bits across a channel, routing, and directing information to the correct
destination within the destination computers operating system demands an
overarching concept of how to organize information delivery. No unique set of rules
satisfies the various constraints communication channels and network organization
place on information transmission. For example, random access issues in Ethernet are
not present in wide-area networks such as the Internet. A protocol is a set of rules
that governs how information is delivered. For example, to use the telephone
network, the protocol is to pick up the phone, listen for a dial tone, dial a number
having a specific number of digits, wait for the phone to ring, and say hello. In radio,
the station uses amplitude or frequency modulation with a specific carrier frequency
and transmission bandwidth, and you know to turn on the radio and tune in the
station. In technical terms, no one protocol or set of protocols can be used for any
communication situation. Be that as it may, communication engineers have found that
a common thread runs through the organization of the various protocols. This grand
design of information transmission organization runs through all modern networks
today.
What has been defined as a networking standard is a layered, hierarchical protocol
organization. As shown in Figure 6.30 (Protocol Picture), protocols are organized by
function and level of detail.
Figure 6.30 Protocol Picture
Protocols are organized according to the level of detail required for information transmission. Protocols at
the lower levels (shown toward the bottom) concern reliable bit transmission. Higher level protocols
concern how bits are organized to represent information, what kind of information is defined by bit
sequences, what software needs the information, and how the information is to be interpreted. Bodies
325
such as the IEEE (Institute for E lectronics and Electrical Engineers) and the ISO (International Standards
Organization) define standards such as this. Despite being a standard, it does not constrain protocol
implementation so much that innovation and competitive individuality are ruled out.
Segregation of information transmission, manipulation, and interpretation into these
categories directly afects how communication systems are organized, and what role(s)
software systems fulfll. Although not thought about in this way in earlier times, this
organizational structure governs the way communication engineers think about all
communication systems, from radio to the Internet.
Exercise 6.37.1
How do the various aspects of establishing and maintaining a telephone conversation
ft into this layered protocol organization?
We now explicitly state whether we are working in the physical layer (signal set design,
for example), the data link layer (source and channel coding), or any other layer. IP
abbreviates Internet protocol, and governs gateways (how information is transmitted
between networks having different internal organizations). TCP (transmission control
protocol) governs how packets are transmitted through a wide-area network such as
the Internet. Telnet is a protocol that concerns how a person at one computer logs on
to another computer across a network. A moderately high level protocol such as
telnet, is not concerned with what data links (wireline or wireless) might have been
used by the network or how packets are routed. Rather, it establishes connections
between computers and directs each byte (presumed to represent a typed character)
to the appropriate operation system component at each end. It is not concerned with
what the characters mean or what programs the person is typing to. That aspect of
information transmission is left to protocols at higher layers.
Recently, an important set of protocols created the World Wide Web. These protocols
exist independently of the Internet. The Internet insures that messages are
transmitted efciently and intact; the Internet is not concerned (to date) with what
messages contain. HTTP (hypertext transfer protocol) frame what messages contain
and what should be done with the data. The extremely rapid development of the Web
on top of an essentially stagnant Internet is but one example of the power of
organizing how information transmission occurs without overly constraining the
details.
6.38 Information Communication Problems
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Problem 6.1: Signals on Transmission Lines
A modulated signal needs to be sent over a transmission line having a characteristic
impedance of Z0 = 50 (Ω) . So that the signal does not interfere with signals others may
be transmitting, it must be bandpass filtered so that its bandwidth is 1 MHz and
centered at 3.5 MHz. The filter's gain should be one in magnitude. An op-amp filter
(Figure 6.31) is proposed.
326
Figure 6.31
1. What is the transfer function between the input voltage and the voltage across
the transmission line?
2. Find values for the resistors and capacitors so that design goals are met.
Problem 6.2: Noise in AM Systems
The signal
emerging from an AM communication system consists of two parts: the message
signal, s (t), and additive noise. The plot (Figure 6.32) shows the message spectrum S (f)
and noise power spectrum PN (f) . The noise power spectrum lies completely within
the signal's band, and has a constant value there of
Figure 6.32
1. What is the message signal's power? What is the signal-to-noise ratio?
2. Because the power in the message decreases with frequency, the signal-to-noise
ratio is not constant within subbands. What is the signal-to-noise ratio in the
upper half of the frequency band?
3. A clever 241 student suggests filtering the message before the transmitter
modulates it so that the signal spectrum is balanced (constant) across frequency.
Realizing that this filtering affects the message signal, the student realizes that
the receiver must also compensate for the message to arrive intact. Draw a block
diagram of this communication system. How does this system's signal-to-noise
ratio compare with that of the usual AM radio?
Problem 6.3: Complementary Filters
Complementary filters usually have "opposite" filtering characteristics (like a lowpass
and a highpass) and have transfer functions that add to one. Mathematically, H1(f) and
H2(f) are complementary if
H1(f)+ H2(f)=1
327
We can use complementary filters to separate a signal into two parts by passing it
through each filter. Each output can then be transmitted separately and the original
signal reconstructed at the receiver. Let's assume the message is bandlimited to W Hz
and that
1. What circuits would be used to produce the complementary filters?
2. Sketch a block diagram for a communication system (transmitter and receiver)
that employs complementary signal transmission to send a message m (t).
3. What is the receiver's signal-to-noise ratio? How does it compare to the standard
system that sends the signal by simple amplitude modulation?
Problem 6.4: Phase Modulation
A message signal m (t) phase modulates a carrier if the transmitted signal equals
where Ï•d is known as the phase deviation. In this problem, the phase deviation is
small. As with all analog modulation schemes, assume that |m (t) | < 1, the message is
bandlimited to W Hz, and the carrier frequency fc is much larger than W .
1. What is the transmission bandwidth?
2. Find a receiver for this modulation scheme.
3. What is the signal-to-noise ratio of the received signal?
HINT: Use the facts that
for small x.
Problem 6.5: Digital Amplitude Modulation
Two ELEC 241 students disagree about a homework problem. The issue concerns the
discrete-time signal s(n) cos(2πf0n) , where the signal s(n) has no special characteristics
and the modulation frequency f0 is known. Sammy says that he can recover s(n) from
its amplitude-modulated version by the same approach used in analog
communications. Samantha says that approach won't work.
1. What is the spectrum of the modulated signal?
2. Who is correct? Why?
3. The teaching assistant does not want to take sides. He tells them that if s(n)
cos(2πf0n) and s(n) sin(2πf0n)were both available, s(n) can be recovered. What
does he have in mind?
Problem 6.6: Anti-Jamming
One way for someone to keep people from receiving an AM transmission is to
transmit noise at the same carrier frequency. Thus, if the carrier frequency is fc so that
the transmitted signal is
the jammer would transmit
328
The noise n(t) has a constant power density spectrum over the bandwidth of the
message m(t). The channel adds white noise of spectral height
a) What would be the output of a traditional AM receiver tuned to the carrier
frequency fc?
b) RU Electronics proposes to counteract jamming by using a different modulation
scheme. The scheme's transmitted signal has the form
where c(t) is a periodic carrier signal (period
) having the indicated waveform (Figure 6.33). What is the spectrum of the transmitted
signal with the proposed scheme? Assume the message bandwidth W is much less
than the fundamental carrier frequency fc.
c) The jammer, unaware of the change, is transmitting with a carrier frequency of fc,
while the receiver tunes a standard AM receiver to a harmonic of the carrier
frequency. What is the signal-to-noise ratio of the receiver tuned to the harmonic
having the largest power that does not contain the jammer?
Figure 6.33
Problem 6.7: Secret Communications
A system for hiding AM transmissions has the transmitter randomly switching
between two carrier frequencies f1 and f2. "Random switching" means that one carrier
frequency is used for some period of time, switches to the other for some other
period of time, back to the first, etc. The receiver knows what the carrier frequencies
are but not when carrier frequency switches occur. Consequently, the receiver must
be designed to receive the transmissions regardless of which carrier frequency is
used. Assume the message signal has bandwidth W. The channel adds white noise of
spectral height
1. How different should the carrier frequencies be so that the message could be
received?
2. What receiver would you design?
3. What signal-to-noise ratio for the demodulated signal does your receiver yield?
Problem 6.8: AM Stereo
329
Stereophonic radio transmits two signals simultaneously that correspond to what
comes out of the left and right speakers of the receiving radio. While FM stereo is
commonplace, AM stereo is not, but is much simpler to understand and analyze. An
amazing aspect of AM stereo is that both signals are transmitted within the same
bandwidth as used to transmit just one. Assume the left and right signals are
bandlimited to W Hz.
x (t)= A (1 + ml (t)) cos (2πfct)+ Amr (t) sin (2πfct)
1. Find the Fourier transform of x(t). What is the transmission bandwidth and how
does it compare with that of standard AM?
2. Let us use a coherent demodulator as the receiver, shown in Figure 6.34. Show
that this receiver indeed works: It produces the left and right signals separately.
3. Assume the channel adds white noise to the transmitted signal. Find the signal-tonoise ratio of each signal.
Figure 6.34
Problem 6.9: A Novel Communication System
A clever system designer claims that the depicted transmitter (Figure 6.35) has,
despite its complexity, advantages over the usual amplitude modulation system. The
message signal m(t) is bandlimited to W Hz, and the carrier frequency fc» W . The
channel attenuates the transmitted signal x(t) and adds white noise of spectral height
Figure 6.35
The transfer function H(f) is given by
1.
2.
3.
4.
Find an expression for the spectrum of x(t). Sketch your answer.
Show that the usual coherent receiver demodulates this signal.
Find the signal-to-noise ratio that results when this receiver is used.
Find a superior receiver (one that yields a better signal-to-noise ratio), and
analyze its performance.
330
Problem 6.10: Multi-Tone Digital Communication
In a so-called multi-tone system, several bits are gathered together and transmitted
simultaneously on different carrier frequencies during a T second interval. For
example, B bits would be transmitted according to
(6.61)
Here, f0 is the frequency offset for each bit and it is harmonically related to the bit
interval T. The value of bk is either −1 or +1.
1. Find a receiver for this transmission scheme.
2. An ELEC 241 almuni likes digital systems so much that he decides to produce a
discrete-time version. He samples the received signal
How should N be related to B, the number of simultaneously transmitted bits?
3. The alumni wants to find a simple form for the receiver so that his software
implementation runs as efficiently as possible. How would you recommend he
implement the receiver?
Problem 6.11: City Radio Channels
In addition to additive white noise, metropolitan cellular radio channels also contain
multipath: the attenuated signal and a delayed, further attenuated signal are received
superimposed. As shown in Figure 6.36, multipath occurs because the buildings reflect
the signal and the reflected path length between transmitter and receiver is longer
than the direct path.
Figure 6.36
Assume that the length of the direct path is d meters and the reflected path is 1.5
times as long. What is the model for the channel, including the multipath and the
additive noise?
Assume d is 1 km. Find and sketch the magnitude of the transfer function for the
multipath component of the channel. How would you characterize this transfer
function?
331
Would the multipath affect AM radio? If not, why not; if so, how so? Would analog
cellular telephone, which operates at much higher carrier frequencies (800 MHz vs. 1
MHz for radio), be affected or not? Analog cellular telephone uses amplitude
modulation to transmit voice.
How would the usual AM receiver be modified to minimize multipath effects? Express
your modified receiver as a block diagram.
Problem 6.12: Downlink Signal Sets
In digital cellular telephone systems, the base station (transmitter) needs to relay
different voice signals to several telephones at the same time. Rather than send
signals at different frequencies, a clever Rice engineer suggests using a different signal
set for each data stream. For example, for two simultaneous data streams, she
suggests BPSK signal sets that have the depicted basic signals (Figure 6.37).
Figure 6.37
Thus, bits are represented in data stream 1 by s1(t) and −s1(t) and in data stream 2 by
s2(t) and −s2(t), each of which are modulated by 900 MHz carrier. The transmitter
sends the two data streams so that their bit intervals align. Each receiver uses a
matched filter for its receiver. The requirement is that each receiver not receive the
other's bit stream.
1. What is the block diagram describing the proposed system?
2. What is the transmission bandwidth required by the proposed system?
3. Will the proposal work? Does the fact that the two data streams are transmitted
in the same bandwidth at the same time mean that each receiver's performance
is affected? Can each bit stream be received without interference from the other?
Problem 6.13: Mixed Analog and Digital Transmission
A signal m(t) is transmitted using amplitude modulation in the usual way. The signal
has bandwidth W Hz, and the carrier frequency is fc. In addition to sending this analog
signal, the transmitter also wants to send ASCII text in an auxiliary band that lies
slightly above the analog transmission band. Using an 8-bit representation of the
characters and a simple baseband BPSK signal set (the constant signal +1 corresponds
to a 0, the constant -1 to a 1), the data signal d(t) representing the text is transmitted
as the same time as the analog signal m(t). The transmission signal spectrum is as
shown (Figure 6.38), and has a total bandwidth B.
332
Figure 6.38
1. Write an expression for the time-domain version of the transmitted signal in
terms of m(t) and the digital signal d(t).
2. What is the maximum datarate the scheme can provide in terms of the available
bandwidth?
3. Find a receiver that yields both the analog signal and the bit stream.
Problem 6.14: Digital Stereo
Just as with analog communication, it should be possible to send two signals
simultaneously over a digital channel. Assume you have two CD-quality signals (each
sampled at 44.1 kHz with 16 bits/sample). One suggested transmission scheme is to
use a quadrature BPSK scheme. If b
transmitted signal has the form
(1)
(n) and b
(2)
(n) each represent a bit stream, the
(1)
(2)
where p(t) is a unit-amplitude pulse having duration T and b (n), b (n) equal either
+1 or -1 according to the bit being transmitted for each signal. The channel adds white
noise and attenuates the transmitted signal.
1. What value would you choose for the carrier frequency fc?
2. What is the transmission bandwidth?
3. What receiver would you design that would yield both bit streams?
Problem 6.15: Digital and Analog Speech Communication
Suppose we transmit speech signals over comparable digital and analog channels. We
want to compare the resulting quality of the received signals. Assume the transmitters
use the same power, and the channels introduce the same attenuation and additive
white noise. Assume the speech signal has a 4 kHz bandwidth and, in the digital case,
is sampled at an 8 kHz rate with eight-bit A/D conversion. Assume simple binary
source coding and a modulated BPSK transmission scheme.
1. What is the transmission bandwidth of the analog (AM) and digital schemes?
2. Assume the speech signal's amplitude has a magnitude less than one. What is
maximum amplitude quantization error introduced by the A/D converter?
3. In the digital case, each bit in quantized speech sample is received in error with
probability pe that depends on signal-to-noise ratio
333
However, errors in each bit have a different impact on the error in N0 the
reconstructed speech sample. Find the mean-squared error between the
transmitted and received amplitude.
4. In the digital case, the recovered speech signal can be considered to have two
noise sources added to each sample's true value: One is the A/D amplitude
quantization noise and the second is due to channel errors. Because these are
separate, the total noise power equals the sum of these two. What is the signalto-noise ratio of the received speech signal as a function of pe?
5. Compute and plot the received signal's signal-to-noise ratio for the two
transmission schemes as a function of channel signal-to-noise ratio.
6. Compare and evaluate these systems.
Problem 6.16: Source Compression
Consider the following 5-letter source.
Letter
Probability
a
0.5
b
0.25
c
0.125
d
0.0625
e
0.0625
1. Find this source's entropy.
2. Show that the simple binary coding is inefcient.
3. Find an unequal-length codebook for this sequence that satisfes the Source
Coding Theorem. Does your code achieve the entropy limit?
4. How much more efcient is this code than the simple binary code?
Problem 6.17: Source Compression
Consider the following 5-letter source.
334
Letter
Probability
a
0.4
b
0.2
c
0.15
d
0.15
e
0.1
1. Find this source's entropy.
2. Show that the simple binary coding is inefficient.
3. Find the Huffman code for this source. What is its average code length?
Problem 6.18: Speech Compression
When we sample a signal, such as speech, we quantize the signal's amplitude to a set
b
of integers. For a b-bit converter, signal amplitudes are represented by 2 integers.
Although these integers could be represented by a binary code for digital
transmission, we should consider whether a Huffman coding would be more efficient.
1. Load into Matlab the segment of speech contained in
y.mat
. Its sampled values
lie in the interval (-1, 1). To simulate a 3-bit converter, we use Matlab's round
function to create quantized amplitudes corresponding to the integers [0l23456
7J.
1. y quant = round(3.5*y + 3.5) ; Find the relative frequency of occurrence of
quantized amplitude values. The following Matlab program computes the
number of times each quantized value occurs.
2. for n=0:7; count(n+l) = sum(y quant == n); end; Find the entropy of this
source.
2. Find the Hufman code for this source. How would you characterize this source
code in words?
3. How many fewer bits would be used in transmitting this speech segment with
your Hufman code in comparison to simple binary coding?
Problem 6.19: Digital Communication
In a digital cellular system, a signal bandlimited to 5 kHz is sampled with a two-bit A/D
converter at its Nyquist frequency. The sample values are found to have the shown
relative frequencies.
335
Sample Value
Probability
0
0.15
1
0.35
2
0.3
3
0.2
We send the bit stream consisting of Hufman-coded samples using one of the two
depicted signal sets (Figure 6.39).
Figure 6.39
1. a) What is the datarate of the compressed source?
2. b) Which choice of signal set maximizes the communication system's
performance?
3. c) With no error-correcting coding, what signal-to-noise ratio would be needed for
your chosen signal set to guarantee that the bit error probability will not exceed
−3
10 ? If the receiver moves twice as far from the transmitter (relative to the
distance at which the 10
change?
−3
error rate was obtained), how does the performance
Problem 6.20: Signal Compression
Letters drawn from a four-symbol alphabet have the indicated probabilities.
Letter
Probability
a
1/3
b
1/3
c
1/4
d
1/12
336
Table 6.8
1. What is the average number of bits necessary to represent this alphabet?
2. Using a simple binary code for this alphabet, a two-bit block of data bits naturally
emerges. Find an error correcting code for two-bit data blocks that corrects all
single-bit errors.
3. How would you modify your code so that the probability of the letter a being
confused with the letter d is minimized? If so, what is your new code; if not,
demonstrate that this goal cannot be achieved.
Problem 6.21: Universal Product Code
The Universal Product Code (UPC), often known as a bar code, labels virtually every
sold good. An example (Figure 6.40) of a portion of the code is shown.
Figure 6.40
Here a sequence of black and white bars, each having width d, presents an 11-digit
number (consisting of decimal digits) that uniquely identifes the product. In retail
stores, laser scanners read this code, and after accessing a database of prices, enter
the price into the cash register.
1. How many bars must be used to represent a single digit?
2. A complication of the laser scanning system is that the bar code must be read
either forwards or backwards. Now how many bars are needed to represent each
digit?
3. What is the probability that the 11-digit code is read correctly if the probability of
reading a single bit incorrectly is pe?
4. How many error correcting bars would need to be present so that any single bar
error occurring in the 11-digit code can be corrected?
Problem 6.22: Error Correcting Codes
A code maps pairs of information bits into codewords of length 5 as follows.
337
Data
Codeword
00
00000
01
01101
10
10111
11
11010
1. What is this code's efficiency?
2. Find the generator matrix G and parity-check matrix H for this code.
3. Give the decoding table for this code. How many patterns of 1, 2, and 3 errors are
correctly decoded?
4. What is the block error probability (the probability of any number of errors
occurring in the decoded codeword)?
Problem 6.23: Digital Communication
A digital source produces sequences of nine letters with the following probabilities.
letter
a
b
c
d
e
f
g
h
i
probability
(a). Find a Huffman code that compresses this source. How does the resulting code
compare with the best possible code?
(b). A clever engineer proposes the following (6,3) code to correct errors after
transmission through a digital channel.
c1=d1
c2=d2
c3=d3
c6=d1
What is the error correction capability of this code?
(c). The channel's bit error probability is 1/8. What kind of code should be used to
transmit data over this channel?
Problem 6.24: Overly Designed Error Correction Codes
An Aggie engineer wants not only to have codewords for his data, but also to hide the
information from Rice engineers (no fear of the UT engineers). He decides to
represent 3-bit data with 6-bit codewords in which none of the data bits appear
explicitly.
338
c1 = d1 ⊕ d2
c4 = d1 ⊕ d2 ⊕ d3
c2 = d2 ⊕ d3
c5 = d1 ⊕ d2
c3 = d1 ⊕ d3
c6 = d1 ⊕ d2 ⊕ d3
1. Find the generator matrix G and parity-check matrix H for this code.
2. Find a 3 × 6 matrix that recovers the data bits from the codeword.
3. What is the error correcting capability of the code?
Problem 6.25: Error Correction?
It is important to realize that when more transmission errors than can be corrected,
error correction algorithms believe that a smaller number of errors have occurred and
correct accordingly. For example, consider a (7,4) Hamming code having the generator
matrix
This code corrects all single-bit error, but if a double bit error occurs, it corrects using
a single-bit error correction approach.
1. How many double-bit errors can occur in a codeword?
2. For each double-bit error pattern, what is the result of channel decoding? Express
your result as a binary error sequence for the data bits.
Problem 6.26: Selective Error Correction
We have found that digital transmission errors occur with a probability that remains
constant no matter how "important" the bit may be. For example, in transmitting
digitized signals, errors occur as frequently for the most significant bit as they do for
the least significant bit. Yet, the former errors have a much larger impact on the
overall signal-to-noise ratio than the latter. Rather than applying error correction to
each sample value, why not concentrate the error correction on the most important
bits? Assume that we sample an 8 kHz signal with an 8-bit A/D converter. We use
single-bit error correction on the most significant four bits and none on the least
significant four. Bits are transmitted using a modulated BPSK signal set over an
additive white noise channel.
1. How many error correction bits must be added to provide single-bit error
correction on the most significant bits?
2. How large must the signal-to-noise ratio of the received signal be to insure
reliable communication?
339
3. Assume that once error correction is applied, only the least significant 4 bits can
be received in error. How much would the output signal-to-noise ratio improve
using this error correction scheme?
Problem 6.27: Compact Disk
Errors occur in reading audio compact disks. Very few errors are due to noise in the
compact disk player; most occur because of dust and scratches on the disk surface.
Because scratches span several bits, a single-bit error is rare; several consecutive bits
in error are much more common. Assume that scratch and dust-induced errors are
four or fewer consecutive bits long. The audio CD standard requires 16-bit, 44.1 kHz
analog-to-digital conversion of each channel of the stereo analog signal.
1. How many error-correction bits are required to correct scratch-induced errors for
each 16-bit sample?
2. Rather than use a code that can correct several errors in a codeword, a clever 241
engineer proposes interleaving consecutive coded samples. As the cartoon
(Figure 6.41) shows, the bits representing coded samples are interpersed before
they are written on the CD. The CD player de-interleaves the coded data, then
performs error-correction. Now, evaluate this proposed scheme with respect to
the non-interleaved one.
Figure 6.41
Problem 6.28: Communication System Design
RU Communication Systems has been asked to design a communication system that
meets the following requirements.
• The baseband message signal has a bandwidth of 10 kHz.
• The RUCS engineers find that the entropy H of the sampled message signal
depends on how many bits b are used in the A/D converter (see table below).
• The signal is to be sent through a noisy channel having a bandwidth of 25 kHz
channel centered at 2 MHz and a signal-to-noise ration within that band of 10 dB.
• Once received, the message signal must have a signal-to-noise ratio of at least 20
dB.
340
b
H
3
2.19
4
3.25
5
4.28
6
5.35
Can these specifcations be met? Justify your answer.
Problem 6.29: HDTV
As HDTV (high-Definition television) was being developed, the FCC restricted this
digital system to use in the same bandwidth (6 MHz) as its analog (AM) counterpart.
HDTV video is sampled on a 1035 × 1840 raster at 30 images per second for each of
the three colors. The least-acceptable picture received by television sets located at an
analog station's broadcast perimeter has a signal-to-noise ratio of about 10 dB.
1. Using signal-to-noise ratio as the criterion, how many bits per sample must be
used to guarantee that a high-quality picture, which achieves a signal-to-noise
ratio of 20 dB, can be received by any HDTV set within the same broadcast
region?
2. Assuming the digital television channel has the same characteristics as an analog
one, how much compression must HDTV systems employ?
Problem 6.30: Digital Cellular Telephones
In designing a digital version of a wireless telephone, you must first consider certain
fundamentals. First of all, the quality of the received signal, as measured by the signalto-noise ratio, must be at least as good as that provided by wireline telephones (30 dB)
and the message bandwidth must be the same as wireline telephone. The signal-tonoise ratio of the allocated wirelss channel, which has a 5 kHz bandwidth, measured
100 meters from the tower is 70 dB. The desired range for a cell is 1 km. Can a digital
cellphone system be designed according to these criteria?
Problem 6.31: Optimal Ethernet Random Access Protocols
Assume a population of N computers want to transmit information on a random
access channel. The access algorithm works as follows.
• Before transmitting, flip a coin that has probability p of coming up heads
• If only one of the N computer's coins comes up heads, its transmission occurs
successfully, and the others must wait until that transmission is complete and
then resume the algorithm.
• If none or more than one head comes up, the N computers will either remain
silent (no heads) or a collision will occur (more than one head). This unsuccessful
transmission situation will be detected by all computers once the signals have
341
propagated the length of the cable, and the algorithm resumes (return to the
beginning).
1. What is the optimal probability to use for flipping the coin? In other words, what
should p be to maximize the probability that exactly one computer transmits?
2. What is the probability of one computer transmitting when this optimal value of p
is used as the number of computers grows to infinity?
3. Using this optimal probability, what is the average number of coin flips that will
be necessary to resolve the access so that one computer successfully transmits?
4. Evaluate this algorithm. Is it realistic? Is it efficient?
Problem 6.32: Repeaters
Because signals attenuate with distance from the transmitter, repeaters are
frequently employed for both analog and digital communication. For example, let's
assume that the transmitter and receiver are D m apart, and a repeater is positioned
halfway between them (Figure 6.42). What the repeater does is amplify its received
signal to exactly cancel the attenuation encountered along the first leg and to retransmit the signal to the ultimate receiver. However, the signal the repeater receives
contains white noise as well as the transmitted signal. The receiver experiences the
same amount of white noise as the repeater.
Figure 6.42
1. What is the block diagram for this system?
2. For an amplitude-modulation communication system, what is the signal-to-noise
ratio of the demodulated signal at the receiver? Is this better or worse than the
signal-to-noise ratio when no repeater is present?
3. For digital communication, we must consider the system's capacity. Is the capacity
larger with the repeater system than without it? If so, when; if not, why not?
Problem 6.33: Designing a Speech Communication System
We want to examine both analog and digital communication alternatives for a
dedicated speech transmission system. Assume the speech signal has a 5 kHz
bandwidth. The wireless link between transmitter and receiver is such that 200 watts
of power can be received at a pre-assigned carrier frequency. We have some latitude
in choosing the transmission bandwidth, but the noise power added by the channel
increases with bandwidth with a proportionality constant of 0.1 watt/kHz.
1. Design an analog system for sending speech under this scenario. What is the
received signal-to-noise ratio under these design constraints?
2. How many bits must be used in the A/D converter to achieve the same signal-tonoise ratio?
3. Is the bandwidth required by the digital channel to send the samples without
error greater or smaller than the analog bandwidth?
342
Problem 6.34: Digital vs. Analog
You are the Chairman/Chairwoman of the FCC. The frequency band 3 MHz to 3.5 MHz
has been allocated for a new "high-quality" AM band. Each station licensed for this
band will transmit signals having a bandwidth of 10 kHz, twice the message bandwidth
of what current stations can send.
1. How many stations can be allocated to this band and with what carrier
frequencies?
2. Looking ahead, conversion to digital transmission is not far in the future. The
characteristics of the new digital radio system need to be established and you are
the boss! Detail the characteristics of the analog-to-digital converter that must be
used to prevent aliasing and ensure a signal-to-noise ratio of 25 dB.
3. Without employing compression, how many digital radio stations could be
allocated to the band if each station used BPSK modulation? Evaluate this design
approach.
6.39 Solutions to Exercises in Chapter 6
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 6.3.1
In both cases, the answer depends less on geometry than on material properties. For
coaxial cable,
For twisted pair,
Solution to Exercise 6.3.2
You can find these frequencies from the spectrum allocation chart (Frequency
Allocations (Page 354)). Light in the middle of the visible band has a wavelength of
about 600 nm, which corresponds to a frequency of 5 × 1014Hz . Cable television
transmits within the same frequency band as broadcast television (about 200 MHz
8
or 2 × 10 Hz ). Thus, the visible electromagnetic frequencies are over six orders of
magnitude higher!
Solution to Exercise 6.3.3
As frequency increases,
and
343
In this high-frequency region,
(6.65)
Thus, the attenuation (space) constant equals the real part of this expression, and
equals
Solution to Exercise 6.4.1
As shown previously (Modulated Communication (Page 277)), voltages and currents in
a wireline channel, which is modeled as a transmission line having resistance,
capacitance and inductance, decay exponentially with distance. The inverse-square
law governs free-space propagation because such propagation is lossless, with the
inverse-square law a consequence of the conservation of power. The exponential
decay of wireline channels occurs because they have losses and some filtering.
Solution to Exercise 6.5.1
Figure 6.43 Figure 6.43
2
2
2
Use the Pythagorean Theorem, (h + R) = R + d , where h is the antenna height, d is
the distance from the top of the earth to a tangency point with the earth's surface,
and R the earth's radius. The line-of-sight distance between two earth-based antennae
equals
As the earth's radius is much larger than the antenna height, we have to a good
approximation that
If one antenna is at ground elevation, say h2 =0 , the other antenna's range is
344
Solution to Exercise 6.5.2
As frequency decreases, wavelength increases and can approach the distance
between the earth's surface and the ionosphere. Assuming a distance between the
two of 80 km, the relation λf = c gives a corresponding frequency of 3.75 kHz. Such low
carrier frequencies would be limited to low bandwidth analog communication and to
low datarate digital communications. The US Navy did use such a communication
scheme to reach all of its submarines at once.
Solution to Exercise 6.7.1
Transmission to the satellite, known as the uplink, encounters inverse-square law
power losses. Reflecting of the ionosphere not only encounters the same loss, but
twice. Refection is the same as transmitting exactly what arrives, which means that the
total loss is the product of the uplink and downlink losses. The geosynchronous orbit
lies at an altitude of 35700km. The ionosphere begins at an altitude of about 50 km.
−8
The amplitude loss in the satellite case is proportional to 2.8 × 10 ; for Marconi, it was
proportional to 4.4 × 10
−10
. Marconi was very lucky.
Solution to Exercise 6.8.1
If the interferer's spectrum does not overlap that of our communications channel - the
interferer is out-of-band we need only use a bandpass filter that selects our
transmission band and removes other portions of the spectrum.
Solution to Exercise 6.9.1
The additive-noise channel is not linear because it does not have the zero-input-zerooutput property (even though we might transmit nothing, the receiver's input consists
of noise).
Solution to Exercise 6.11.1
The signal-related portion of the transmitted spectrum is given by
Multiplying at the receiver by the carrier shifts this spectrum to fc and to −fc, and scales the
result by half.
(6.67)
The signal components centered at twice the carrier frequency are removed by the
lowpass filter, while the baseband signal M(f) emerges.
Solution to Exercise 6.12.1
The key here is that the two spectra M (f − fc), M (f + fc) do not overlap because we have
assumed that the carrier frequency fc is much greater than the signal's highest
345
frequency. Consequently, the term M (f − fc) M (f + fc) normally obtained in computing
the magnitude-squared equals zero.
Solution to Exercise 6.12.2
Separation is 2W. Commercial AM signal bandwidth is 5 kHz. Speech is well contained
in this bandwidth, much better than in the telephone!
Solution to Exercise 6.13.1
Solution to Exercise 6.14.1
k =4.
Solution to Exercise 6.14.2
Solution to Exercise 6.14.3
The harmonic distortion is 10%.
Solution to Exercise 6.14.4
Twice the baseband bandwidth because both positive and negative frequencies are
shifted to the carrier by the modulation: 3R.
Solution to Exercise 6.16.1
In BPSK, the signals are negatives of each other: s1 (t)= −s0 (t). Consequently, the
output of each multiplier-integrator combination is the negative of the other.
Choosing the largest therefore amounts to choosing which one is positive. We only
need to calculate one of these. If it is positive, we are done. If it is negative, we choose
the other signal.
Solution to Exercise 6.16.2
The matched filter outputs are
because the sinusoid has less power than a pulse having the same amplitude.
Solution to Exercise 6.17.1
2
The noise-free integrator outputs differ by αA T , the factor of two smaller value than
in the baseband case arising because the sinusoidal signals have less energy for the
same amplitude. Stated in terms of Eb, the difference equals 2αEb just as in the
baseband case.
Solution to Exercise 6.18.1
The noise-free integrator output difference now equals
346
The noise power remains the same as in the BPSK case, which from the probability of
error equation (6.46) yields
Solution to Exercise 6.20.1
Equally likely symbols each have a probability of
Thus,
To prove that this is the maximum-entropy probability assignment, we must explicitly
take into account that probabilities sum to one. Focus on a particular symbol, say the
first. Pr [a0] appears twice in the entropy formula: the terms Pr[a0] log2Pr[a0]and (1
−Pr[a0]+ ··· + Pr[aK−2]) log2(1 −Pr [a0]+ ··· + Pr [aK−2]) . The derivative with respect to this
probability (and all the others) must be zero. The derivative equals log 2Pr [a0] − log2 (1
−Pr [a0]+ ··· + Pr[aK−2]), and all other derivatives have the same form (just substitute
your letter's index). Thus, each probability must equal the others, and we are done.
For the minimum entropy answer, one term is 1log21=0 , and the others are 0log20 ,
which we define to be zero also. The minimum value of entropy is zero.
Solution to Exercise 6.22.1
The Huffman coding tree for the second set of probabilities is identical to that for the
first (Figure 6.18 (Huffman Coding Tree)). The average code length is
The entropy calculation 2 4 5 20 is straightforward:
which equals 1.68 bits.
Solution to Exercise 6.23.1
Solution to Exercise 6.23.2
Because no codeword begins with another's codeword, the first codeword
encountered in a bit stream must be the right one. Note that we must start at the
beginning of the bit stream; jumping into the middle does not guarantee perfect
decoding. The end of one codeword and the beginning of another could be a
codeword, and we would get lost.
Solution to Exercise 6.23.3
347
Consider the bitstream ...0110111... taken from the bitstream 0I10I110I110I111I.... We
would decode the initial part incorrectly, then would synchronize. If we had a fxedlength code (say 00,01,10,11), the situation is much worse. Jumping into the middle
leads to no synchronization at all!
Solution to Exercise 6.25.1
2
2
This question is equivalent to 3pe×(1 − pe)+pe ≤ 1 or 2pe −3pe +1 ≥ 0. Because this is an
upward-going parabola, we need only check where its roots are. Using the quadratic
formula, we find that they are located at
and 1. Consequently in the range
the error rate produced by coding is smaller.
Solution to Exercise 6.26.1
With no coding, the average bit-error probability pe is given by the probability of error
equation (6.47):
With a threefold repetition code, the bit-error probability is given by
where
Plotting this reveals that the increase in bit-error probability out of the channel
because of the energy reduction is not compensated by the repetition coding.
348
Figure 6.44 Error Probability with and without (3,1) Repetition Coding
Solution to Exercise 6.27.1
In binary arithmetic (see 6.27 No Title Provided), adding 0 to a binary value results in
that binary value while adding 1 results in the opposite binary value.
Solution to Exercise 6.27.2
dmin=2n +1
Solution to Exercise 6.28.1
When we multiply the parity-check matrix times any codeword equal to a column of G,
the result consists of the sum of an entry from the lower portion of G and itself that,
by the laws of binary arithmetic, is always zero.
Because the code is linear - sum of any two codewords is a codeword - we can
generate all codewords as sums of columns of G. Since multiplying by H is also linear,
Hc = 0.
Solution to Exercise 6.28.2
In binary arithmetic see this table ("Error Correction" <http://legacy.cnx.org/content/
m0095/latest/#table1>), adding 0 to a binary value results in that binary value while
adding 1 results in the opposite binary value.
Solution to Exercise 6.28.3
349
The probability of a single-bit error in a length-N block is Npe(1 − pe)
N-1
and a triple-bit
error has probability
For the first to be greater than the second, we must have
For N =7, pe< 0.31.
Solution to Exercise 6.29.1
In a length-N block, N single-bit and
double-bit errors can occur. The number of non-zero vectors resulting from
must equal or exceed the sum of these two numbers.
(6.68)
The first two solutions that attain equality are (5,1) and (90,78) codes. However, no
perfect code exists other than the single-bit error correcting Hamming code. (Perfect
codes satisfy relations like (6.68) with equality.)
Solution to Exercise 6.31.1
To convert to bits/second, we divide the capacity stated in bits/transmission by the bit
interval duration T.
Solution to Exercise 6.33.1
The network entry point is the telephone handset, which connects you to the nearest
station. Dialing the telephone number informs the network of who will be the
message recipient. The telephone system forms an electrical circuit between your
handset and your friend's handset. Your friend receives the message via the same
device the handset that served as the network entry point.
Solution to Exercise 6.36.1
The transmitting op-amp sees a load or
where N is the number of transceivers other than this one attached to the coaxial
cable. The transfer function to some other transceiver's receiver circuit is Rout divided
by this load.
Solution to Exercise 6.36.2
The worst-case situation occurs when one computer begins to transmit just before the
other's packet arrives. Transmitters must sense a collision before packet transmission
350
ends. The time taken for one computer's packet to travel the Ethernet's length and for
the other computer's transmission to arrive equals the round-trip, not one-way,
propagation time.
Solution to Exercise 6.36.3
The cable must be a factor of ten shorter: It cannot exceed 100 m. Different minimum
packet sizes means different packet formats, making connecting old and new systems
together more complex than need be.
Solution to Exercise 6.37.1
When you pick up the telephone, you initiate a dialog with your network interface by
dialing the number. The network looks up where the destination corresponding to
that number is located, and routes the call accordingly. The route remains fixed as
long as the call persists. What you say amounts to high-level protocol while
establishing the connection and maintaining it corresponds to low-level protocol.
351
Chapter 7 Appendix
7.1 Decibels
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The decibel scale expresses amplitudes and power values logarithmically. The
definitions for these differ, but are consistent with each other.
(7.1)
Here power (s0) and amplitude (s0) represent a reference power and amplitude,
respectively. Quantifying power or amplitude in decibels essentially means that we are
comparing quantities to a standard or that we want to express how they changed. You
will hear statements like "The signal went down by 3 dB" and "The filter's gain in the
stopband is -60" (Decibels is abbreviated dB.).
Exercise 7.1.1
The prefix "deci" implies a tenth; a decibel is a tenth of a Bel. Who is this measure
named for?
The consistency of these two definitions arises because power is proportional to the
square of amplitude:
(7.2)
Plugging this expression into the Definition for decibels, we find that
(7.3)
Because of this consistency, stating relative change in terms of decibels is
unambiguous. A factor of 10 increase in amplitude corresponds to a 20 dB increase in
both amplitude and power!
352
Power Ratio
dB
1
0
1.5
2
3
5
4
6
5
7
8
9
10
10
0.1
-10
Table 7.1 Decibel table Common values for the decibel. The decibel
values for all but the powers of ten are approximate, but are accurate
to a decimal place.
The accompanying table provides "nice" decibel values. Converting decibel values back
and forth is fun, and tests your ability to think of decibel values as sums and/or
differences of the well-known values and of ratios as products and/or quotients. This
conversion rests on the logarithmic nature of the decibel scale.
For example, to find the decibel value for
we halve the decibel value for 2; 26 dB equals 10 + 10 + 6 dB that corresponds to a
ratio of 10 × 10 × 4 = 400. Decibel quantities add; ratio values multiply.
One reason decibels are used so much is the frequency-domain input-output relation
for linear systems: Y (f)= X (f) H (f) . Because the transfer function multiplies the input
signal's spectrum, to find the output amplitude at a given frequency we simply add the
filter's gain in decibels (relative to a reference of one) to the input amplitude at that
frequency. This calculation is one reason that we plot transfer function magnitude on
a logarithmic vertical scale expressed in decibels.
7.2 Permutations and Combinations
7.2.1 Permutations and Combinations
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
The lottery "game" consists of picking k numbers from a pool of n. For example, you
select 6 numbers out of 60. To win, the order in which you pick the numbers doesn't
matter; you only have to choose the right set of 6 numbers. The chances of winning
equal the number of different length-k sequences that can be chosen. A related, but
353
different, problem is selecting the batting lineup for a baseball team. Now the order
matters, and many more choices are possible than when order does not matter.
Answering such questions occurs in many applications beyond games. In digital
communications, for example, you might ask how many possible double-bit errors can
occur in a codeword. Numbering the bit positions from 1 to N, the answer is the same
as the lottery problem with k = 6. Solving these kind of problems amounts to
understanding permutations -the number of ways of choosing things when order
matters as in baseball lineups -and combinations -the number of ways of choosing
things when order does not matter as in lotteries and bit errors.
Calculating permutations is the easiest. If we are to pick k numbers from a pool of n,
we have n choices for the first one. For the second choice, we have n − 1. The number
of length-two ordered sequences is therefore be n (n − 1). Continuing to choose until
we make k choices means the number of permutations is n (n − 1) (n − 2) ... (n − k + 1).
This result can be written in terms of factorials as
with n!= n (n − 1) (n − 2) ... 1. For mathematical convenience, we define 0! = 1.
When order does not matter, the number of combinations equals the number of
permutations divided by the number of orderings. The number of ways a pool of k
things can be ordered equals k!. Thus, once we choose the nine starters for our
baseball game, we have 9! = 362, 880 different lineups! The symbol for the
combination of k things drawn from a pool of n is
and equals
Exercise 7.2.1
What are the chances of winning the lottery? Assume you pick 6 numbers from the
numbers 1-60. Combinatorials occur in interesting places. For example, Newton
derived that the n-th power of a sum
Exercise 7.2.2
What does the sum of binomial coefficients equal? In other words, what is
A related problem is calculating the probability that any two bits are in error in a
length-n codeword when p is the probability of any bit being in error. The probability
2
n−2
of any particular two-bit error sequence is p (1 − p)
. The probability of a two-bit
error occurring anywhere equals this probability times the number of combinations:
354
Note that the probability that zero or one or two, etc. errors occurring must be one; in
other words, something must happen to the codeword! That means that we must
have
Can you prove this?
7.3 Frequency Allocations
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
To prevent radio stations from transmitting signals "on top of each other," the United
States and other national governments in the 1930s began regulating the carrier
frequencies and power outputs stations could use. With increased use of the radio
spectrum for both public and private use, this regulation has become increasingly
important. This is the so-called Frequency Allocation Chart, which shows what kinds
of broadcasting can occur in which frequency bands. Detailed radio carrier frequency
assignments are much too detailed to present here.
7.4 Solutions to Exercises in Chapter 7
Available under Creative Commons-ShareAlike 4.0 International License (http://creativecommon
s.org/licenses/by-sa/4.0/).
Solution to Exercise 7.1.1
Alexander Graham Bell. He developed it because we seem to perceive physical
quantities like loudness and brightness logarithmically. In other words, percentage,
not absolute differences, matter to us. We use decibels today because common values
are small integers. If we used Bels, they would be decimal fractions, which aren't as
elegant.
Solution to Exercise 7.2.1
Solution to Exercise 7.2.2
n
n
Because of Newton's binomial theorem, the sum equals (1 + 1) =2 .
Download